Content and licensingview original scan buy a printed copy

Geology of the Ballater district. Memoir for 1:50 000 geological map sheet 65E (Scotland)

Bibliographical reference: Smith, C G, Goodman, S, and Robertson, S. 2002. Geology of the Ballater district. Memoir of the British Geological Survey, Sheet 65E (Scotland)

British Geological Survey. In memory of Steven Robertson 1960–2000

London: The Stationery Office 2002. © NERC copyright 2002. First published 2002. ISBN 0 11 884 563 2.

Printed in the UK for The Stationery Office. J103746 C6 06/01

The grid used on the figures is the National Grid taken from the Ordnance Survey maps. (Figure 2) is based on material from Ordnance Survey 1:50 000 scale maps, numbers 37 and 44. © Crown copyright reserved. Ordnance Survey Licence No. GD272191/2002.

(Front cover) Aerial photograph of Lochnagar showing north-facing corrie incised in Lochnagar Granite, and 1050 m erosion surface. (D5236) (Photographer T Bain)

(Rear cover)

Notes

The word ‘district’ in the text refers to the area covered by 1:50 000 Sheet 65E Ballater.

National Grid references are shown in square brackets; all refer to 100 km square NO unless otherwise indicated.

Chemical elements are abbreviated using the standard notation of the Periodic Table of elements.

Four and five-figure numbers prefixed by N and S refer to thin sections lodged in the Scottish sliced rock collection of the British Geological Survey (located at Murchison House).

Acknowledgements

The 1:50 000 Sheet 65E Ballater (Solid Geology) is the result of a mapping programme carried out between 1986 and 1990 by BGS and Aberdeen University, under contract to NERC. This modern survey is built on the first edition of Sheet 65 (1904), and makes full use of the 1:10 560 field maps of G Barrow, E H Cunningham Craig and L W Hinxman. Unpublished geological and geophysical mapping by B Harte, K E Jarvis, W Oldershaw, T Tadesse and F E Tocher was also incorporated.

The ground north-west of the Glen Muick–Glen Girnock watershed and on the south-west side of Glen Clova was mapped by C G Smith who also compiled the map. S Robertson was responsible for mapping the section north of the River Dee, eastwards from Geallaig Hill, and in the area between the Mount Battock Granite and Glen Clova. J R Mendum surveyed the ground delimited by the Glen Doll Fault and the southern edge of the Lochnagar Granite, and D I J Mallick also mapped part of the granite; D Gould covered the Mount Battock Granite and the Dalradian rocks immediately north of it. The Aberdeen University contract mapping was undertaken by S Goodman and A Crane, with A G Leslie (QUB) in a 7 km-wide corridor extending for 20 km from the southern edge of the Ballater Granite to the north-east slopes of Glen Clova.

Air photography at a scale of 1:24 000 was also used, together with geochemical data from the BGS Regional Survey Programme (G-BASE), regional gravity and aero-magnetic information.

The memoir was compiled by C G Smith who also wrote chapters 1, 2, 4, 5, 6 and 16. K Rollin contributed Chapter 3. Chapters 7 to 11, 14 and 15 are composite, encompassing sections by all the authors, and with additional contributions from A G Leslie, A Crane and J R Mendum. Chapters 12 and 13 were compiled by S Goodman, D I J Mallick and C G Smith and chapter 17 was written by S Goodman. The contributions of E R Phillips (chapter 6, 12, 13 and 14) and S Noble (chapter 14) are also acknowledged. The manuscript was edited by A A Jackson.

Figures in this memoir were produced by BGS Drawing Office, Murchison House. The kind co-operation and assistance of numerous landowners and the unstinting hospitality of the people of the district are gratefully acknowledged.

Preface

The Ballater district is an area of outstanding mountain scenery, unique in north-east Scotland, which has continued to attract tourists, hillwalkers and climbers since it was first popularised by Queen Victoria in the mid 19th century. Many of these visitors are especially interested in the natural history of the district and this memoir will help them to comprehend its complex geological evolution. The Ballater district contains one of the classic areas of Scottish geology where George Barrow, in the course of the primary geological survey, first elucidated the concept of metamorphic zones. It is also a key element in linking the Dalradian successions of the central Highlands and north-east Scotland.

For the most part the primary survey did not establish a chronology in either the igneous or metamorphic rocks. In the hundred or so years since then, knowledge and understanding of the Dalradian rocks and the igneous rocks intruded into them has advanced to the extent that a second survey of the solid geology was undertaken between 1986 and 1990; the results are described in this memoir. The survey involved collaboration between the British Geological Survey, the University of Aberdeen and Queen’s University Belfast. The university involvement stemmed from a policy of BGS’s parent body, the Natural Environment Research Council, to encourage university staff to contribute their knowledge about particular areas and their special areas of expertise, and to place their data and information in the public domain. In the case of the Ballater district, the university contribution was based around detailed geophysical and geological surveying of the central part of the district where exposure is poor. The remainder of the district was mapped by BGS. Earlier university work on the Lochnagar Granite has also been incorporated.

Granite and limestones were formerly quarried on Deeside, and the former, together with sand and gravel deposits constitute the principle potential geological resources of the district. Lead was formerly mined on a small scale near Ballater and there are geochemical anomalies of copper, zinc and tin mineralisation elsewhere in the district. However, the main natural resource of the district remains the natural beauty of the landscape, which reflects the underlying geology. It is hoped that this memoir will contribute to the better understanding of the district and help to explain the necessity and the need for conservation of this popular area.

David A Falvey, PhD Director. British Geological Survey, Kingsley Dunham Centre, Keyworth, Nottingham. NG12 5GG

Geology of the Ballater district

This memoir describes the solid geology of the impressively mountainous terrain between upper Deeside and Glen Clova in north-east Scotland. The district is underlain almost entirely by metamorphic and igneous rocks, present in roughly equal proportions. Precursors of the metamorphic rocks were marine sediments and minor volcanic rocks of the Neoproterozoic Dalradian Supergroup. Within the Ballater district there is a near-complete sequence that includes the limestone-shale-quartzite assemblage of the Appin Group, psammites, semipelites and impure limestones in the lower Argyll Group, and a monotonous sequence of metagreywackes, broken only by limestone and quartzite, that characterise the upper Argyll and Southern Highland groups. Metavolcanic rocks, including metabasalts and volcaniclastic turbidites (Green Beds) occur locally in the lower Argyll Group and in the Southern Highland Group. The Dalradian rocks were subjected to polyphase folding and metamorphism during the Caledonian Orogeny. The first and second phase of folding caused widespread inversion of the rocks. A third deformation caused steepening and, in the north-west of the district, produced re-inversion of the strata. Regional (Barrovian) metamorphism, which peaked between the second and third deformations, ranges from staurolite grade in the south-east to sillimanite grade over most of the district, although the extreme north is characterised by Buchan-type metamorphism with the development of andalusite instead of kyanite.

The igneous rocks range in age from Precambrian to Silurian. On the basis of their relationship to tectonism these are divided into four groups: early to syntectonic metadolerites, syntectonic granite, late-tectonic basic, ultrabasic and granitic rocks and post-tectonic granitoids. The last occupy around half of the surface area, and geophysical modelling (based on gravity, aeromagnetic and ground magnetic surveys) indicates that some have far greater subsurface extent. They comprise twelve discrete bodies, dominantly of granite, but including a range from gabbro to highly evolved, lithium-enriched granite. The intrusions are roughly coeval; three new U/Pb zircon-based dates gave a range of 426 to 417 Ma. Estimates of pressure based on thermal metamorphic assemblages, suggest that emplacement took place at a depth of 7 to 8 km and not as a subvolcanic ring complex as previously thought. Minor intrusions including felsites, microdiorites and lamprophyres, mainly in the form of dykes, occur widely throughout the district cutting both metamorphic and major intrusive rocks.

The present-day morphology is largely the result of late Palaeozoic erosion, modified by Miocene tropical weathering and Quaternary glaciation. The district contains minor lead, tungsten and tin mineralisation, and some occurrences of semi-precious stones. The granites represent a significant potential resource of hard-rock aggregate and dimension stone, but the outstanding scenery of the district means that the tourist industry still holds the greatest potential for development.

(Table 1) Geological sequence of the Ballater district.

Solid geology of the Ballater district

This memoir describes the bedrock geology of the impressively mountainous country lying between upper Deeside and upper Glen Clova in north-east Scotland. The area has been favoured by royalty since the mid 19th century when Queen Victoria purchased the Balmoral Estate, the greater part of which lies in the Ballater district. The large upstanding granitic intrusion of Lochnagar. beloved of climbers and walkers and immortalised in verse by Lord Byron, dominates the country.

Lochnagar is one of several granite/diorite masses that were intruded some 420 million years ago into folded and metamorphosed Dalradian metasedimentary and metavolcanic rocks that underlie much of the district. The Dalradian sedimentary rocks were deposited in an ancient ocean in late Precambrian times. They were subsequently intruded by many molten basaltic sheets before being deformed under conditions of great heat and pressure at deep levels in the crust during the Caledonian mountain building which happened between 490 and 400 million years ago.

This memoir outlines the complex nature of the igneous and metamorphic elements that now make up these old mountain roots and form the foundation of this largely heather and pine-clad terrain. These strongly eroded remnants of a once much higher Caledonian mountain chain have been affected by tropical weathering during the Cainozoic (Tertiary) and more recently, by Quaternary glaciation. The erosive effects of glaciation have resulted in formation of comes and U-shaped valleys on the higher ground. whereas the deposition of till and outwash sands and gravels has infilled the main valleys. Minor metalliferous mineralisation is present, but the granites themselves and spring water are recognised as the chief geological resources. However, the main attraction of the area remains its scenic and mountainous nature.

Chapter 1 Introduction

This memoir describes the solid geology of the Ballater district in north-east Scotland, an area of 555 km2 covered by the geological 1:50 000 Series Scotland Sheet 65E Ballater (Figure 1). The district (Figure 2) comprises a deeply dissected plateau that slopes gently from between 700 and 800 m OD in the south-west to 500 and 600 m OD in the north-east. Two mountain masses protrude above this surface; in the west the spectacular bulk of the Lochnagar massif rises to 1155 m OD, the highest point in the area, and on the eastern edge the conical peak of Mount Keen reaches to 939 m OD.

The northern half of the district is drained by the easterly flowing River Dee and its principal tributaries, the rivers Gairn and Muick, and the Water of Tanar. The eastern edge of the plateau is drained by the Water of Lee and the Water of Mark, which are the main headwaters of the River North Esk. In the south-west, the River South Esk flowing to the south-east is the main drainage channel.

There is a marked contrast in valley morphology between the northern and southern parts of the area largely as a result of glacial erosion and the lower elevation of the plateau in the north. This is best illustrated by comparing the valley of the Dee with Glen Clova. For example, around the Dee valley the topography is subdued with gentle slopes and rounded hills giving rise to a wide, open valley ((Plate 1)a). In contrast, Glen Clova is a deeply incised trough, and its upper valley sides are dominated by rocky crags ((Plate 1)b). With its characteristic deep parabolic shape and well-developed corries on its upper slopes, Glen Clova is a classical example of a glacially overdeepened valley. High-level corries occur widely throughout the southern and western parts of the district; probably the most impressive is that forming the precipitous north face of Lochnagar, where the backwall is more than 350 m in height. With the notable exception of Corrie Fee, most of the larger corries contain small lochs. Loch Muick and Loch Lee are larger glacially scoured depressions in their respective glens, the former being over 70 m deep.

In parts of the district the topography reflects the underlying geology. The high tops of Lochnagar and Mount Keen are composed of granite; the plateau south-east of Glen Muick and around Glen Clova is underlain largely by resistant gneissose semipelites and psammites, whereas the low ground around Glen Girnock coincides with the outcrop of softer calcareous schists. The deeply incised nature of the small valleys containing the Moulzie and Kilbo burns result from erosion along the line of the Glen Doll Fault. Likewise, the narrow col between Driesh and Hill of Strone [NO 282 736] is on the Farchal Fault, and the lower reaches of Glen Gairn are aligned parallel to the Lecht Lineament.

Previous research

Barrow and Cunningham Craig (1912) cited a number of references to Upper Deeside in 19th century geological literature. Most of these are within regional accounts, covering many aspects of the rocks, minerals and glacial processes. Although the majority of these have had little impact on current thinking regarding the geological history of the region, they are testimony to the importance of this area in the minds of the early geologists, and some include information on notable mineral occurrences (for example Heddle, 1877).

The area was first surveyed by the Geological Survey in the period 1884 to 1900; the resulting map at a scale of one inch to one mile, geological Sheet 65 Balmoral was published in 1904. The accompanying descriptive memoir (Barrow and Cunningham Craig, 1912) appeared later. Probably the most important result of the primary survey was the recognition by Barrow of three metamorphic zones in this district, characterised by the presence of sillimanite, kyanite and staurolite. Details of these zones, the Barrovian zonal scheme, were first published by Barrow (1893) and, although certain aspects of this paper are not now accepted (for example the metamorphism is regional and not the result of emplacement of the Older Granites), it is still regarded as a milestone in Scottish geology, being the first attempt in the country to establish lines on a map between differing grades of regional metamorphism. Only the ‘approximate outer limit of the sillimanite aureole’ was defined on the published one-inch map, but the distribution of the zones was given by Barrow (1912). Barrow did not recognise the existence of contact metamorphism around the granites and diorites of the Newer Granite suite, but such thermal effects were described later by Macgregor (1929) and Hutchison (1933). The only other paper of note on this area in the first half of the 20th century was by Read (1928) who established a correlation between the Dalradian rocks of the Deeside–Glen Muick area and those of central Perthshire; he also predicted the importance of this area in linking the Dalradian succession of Perthshire with that of north-east Scotland.

In the following 25 years no publication on the geology of the area appeared, although passing reference was made to the district by Macgregor (1948). He included the suggestion that the Lochnagar (granite) mass might possess an annular or ring form, similar to that in the Etive Complex, a concept later developed by Oldershaw (1958, 1974).

Oldershaw’s study marked a revival of interest in the area, which initially focused on a re-examination of Barrow’s pioneering work in Glen Clova. After Harry’s (1958) critical review of the nature of the Older Granites, Chinner (1960, 1961, 1965) investigated some of the chemical controls on the development of the Barrovian index minerals. He looked particularly at the growth of kyanite and sillimanite and the influence of initial ferrous/ferric ratios. In a later more wide-ranging review of metamorphic temperatures and pressures in north-east Scotland, Chinner (1966) concluded that the formation of sillimanite was superimposed on a slightly earlier metamorphism which generated kyanite over much of this area. This sillimanite overprint occurred between the third and fourth phases of deformation (Harte and Johnson, 1969). The assumed real extent of kyanite development was later reduced by the discovery that regional andalusite characterised the rocks lying north-west of Glen Muick (Chinner and Heseltine, 1979). Further mapping in the south-eastern part of the Ballater district (Harte, 1979) established that the Dalradian rocks of upper Glen Esk were part of the Argyll Group in a right-way-up succession separated from the surrounding inverted rocks by a major slide.

In the 1980s the major thrust of research in the Dalradian rocks was directed towards a better understanding of the temperatures and pressures which prevailed during metamorphism; for example McLellan (1985), Baker and Droop (1983), Baker (1985). McLellan (1989) extended the field of research to cover the origin of the two types of migmatites that occur in the area. Later in the decade and into the 1990s the emphasis was mainly on interpreting the major structures, particularly the role of lineaments and shear zones (Goodman, 1994), and increasing use was made of geochemistry and geophysics (Fettes et al., 1986; Tadesse, 1991).

Significantly less time has been devoted to the igneous rocks of the area, although nearly all of the major intrusions have been studied. Rennie (1983) followed up Oldershaw’s earlier work with a mineralogical and geochemical assessment of the Lochnagar Granite; O’Brien (1985) included the Lochnagar, Glen Gairn and Mount Battock intrusions in her nationwide petrogenetic and geochemical study of British Caledonian granites. Geochemical aspects of surface and borehole samples from the Ballater and Mount Battock granites were investigated by Webb and Brown (1984) in a study of heat production in the Eastern Highland granites. The granites around Glen Gairn, with their localised lithium enrichment, have attracted the attention of a number of workers (Orridge, 1961; Hall and Walsh, 1972; Harrison, 1987; Tindle and Webb, 1989). Harrison (1987) also included the Ballater and Mount Battock intrusions in his thesis on the evolution of the Eastern Grampian granites. At the same time Jarvis (1987) compared the petrogenesis and geochemistry of the Glen Doll and Juanjorge diorites. Rb–Sr dating of the Lochnagar (Halliday et al., 1979), Glen Gairn and Mount Battock granites (Harrison and Hutchison, 1987), together with U–Pb studies on Lochnagar (Pigeon and Aftallion, 1978) provided ages ranging from 416 to 404 Ma, confirming that the major granitoids of the district are part of the late Caledonian ‘Newer Granite’ suite. The only recorded study of post-tectonic minor intrusions in the district was by Tocher (1961).

Argentiferous galena and fluorspar mineralisation in the Ballater district, despite its limited extent, has attracted the attention of a number of authors (Wilson and Flett, 1921; Russell, 1937; Lawrie, 1943; Dunham, 1952). The more recent discovery of tungsten-tin-molybdenum-bismuth-silver veins in the zinnwaldite-bearing granite of Glen Gairn has been documented (Harrison, 1987; Webb et al., 1992). Jarvis (1987) attributed high levels of zinc in the Glen Doll Diorite to contamination by Dalradian metasedimentary rocks containing stratabound base metal mineralisation, and the recently published geochemical atlas (British Geological Survey, 1991) specifically draws attention to high levels of copper, lead and zinc in streams draining the Argyll Group to the east of Lochnagar.

Summary of geological history

In broad terms, the Ballater district is underlain by regionally metamorphosed and intrusive igneous rocks in roughly equal proportions (Figure 1). The metamorphic rocks form part of the Dalradian Supergroup, originally a thick prograding (in excess of 25 km) sequence of marine sedimentary and minor basic volcanic rocks that was deposited during Neoproterozoic times (Table 1). The age of sedimentation is constrained by a date of about 750 Ma for syntectonic pegmatites (Piasecki and van Breeman, 1983) which cut the underlying Grampian Group rocks in the Monadhliath Mountains, and the 590 ± 2 Ma date for the Ben Vuirich Granite (Rogers et al., 1989) which was emplaced into Appin Group metasedimentary rocks (Tanner and Leslie, 1994).

In general, progressively younger rocks are encountered as the Dalradian is traversed towards the south-east. North-west of Glen Muick the rocks are right-way-up whereas, to the south-east of this, the succession is for the most part inverted. The oldest Dalradian rocks in the district are quartzites, limestones, calcsilicate-rocks and schistose pelites and semipelites of the Appin Group. The overlying Argyll Group is a more varied sequence of lithologies which includes quartzite, schistose and gneissose psammites, semipelites and pelites, limestone, calcsilicate-rock, calcschist, hornblende schist and amphibolite. The base of the group is marked locally by the presence of a boulder bed, generally considered to be a tillite. Deposition of the lower part of the Argyll Group (Islay and Easdale subgroups) was accompanied by extensional tectonism and sporadic basic volcanism. The higher part of the group (Crinan and Tayvallich subgroups) and the stratigraphically overlying Southern Highland Group are a more monotonous sequence of psammitic, semipelitic and subordinate pelitic metagreywackes which locally includes quartzite, limestone, hornblende schist and volcaniclastic rocks (Green Beds).

The Dalradian rocks were deformed and metamorphosed during the early Palaeozoic Grampian Orogeny (Tanner and Leslie, 1994; Robertson, 1994). The orogeny comprised four episodes of folding and, more locally, the generation of large ductile shear zones. Contemporaneous with the second and third deformation, the rocks underwent prograde regional metamorphism up to amphibolite facies, indicated by the growth of sillimanite, kyanite and andalusite and the development of stromatic and pervasive migmatites.

The district has a long history of intrusive magmatism, commencing with the widespread emplacement of basaltic, doleritic and minor ultramafic sheets prior to the first deformation. A later suite of basic and ultramafic rocks, including the shear-bounded bodies on the Glen Girnock–Glen Muick watershed were emplaced in the Ballater and Crathie areas between the second and third phases of deformation; they are part of a widespread event of basic magmatism in north-east Scotland known as the Younger Basics that were intruded at about 470 Ma (Rogers et al., 1994).

Within the south-eastern part of the Ballater district two generations of syntectonic to late-tectonic granites have been recognised. The older of these, the Rough Craig Granite, was emplaced before the late Cambrian to early Ordovician D2 deformation (Robertson, 1994), whereas the younger Hunt Hill and Cairn Trench granites were intruded during the Ordovician (Robertson, 1991). Collectively they are referred to as Older Granites. More widespread acid magmatism occurred in late Silurian and Devonian times with the emplacement, principally in the north-western half of the district, of large granitoid bodies belonging to the post-tectonic Newer Granite suite. These comprise the Abergeldie, Cul nan Gad, Allt Darrarie, Glen Doll, Juanjorge and Moulzie Burn diorites, and the Ballater, Coilacriech, Glen Gairn, Khantore, Lochnagar and Mount Battock granites. Many of these are multicomponent bodies exhibiting a basic to acid trend with time. Some, such as Glen Doll, are composed largely of dioritic rocks, whereas Abergeldie has a sizeable granodiorite component. U/Pb age determinations have established that the bulk of the Lochnagar Granite was emplaced approximately synchronously, the central (L2) granite and the Abergeldie Diorite being dated at 425 ± 4 Ma. The highly evolved L3 granite has yielded a significantly younger age of 417 ± 1 Ma. These dates compare with previous Rb/Sr determination of 415 ± 3 Ma for Lochnagar L1/ L2 granites (Halliday et al., 1979), 404 ± 6 Ma for Glen Gairn G2/Coilacriech and 416 ± 4 Ma for Mount Battock (Harrison and Hutchinson, 1987). The emplacement of these major intrusions occasioned widespread thermal metamorphism in adjacent Dalradian rocks, particularly those next to the dioritic bodies, with the development of andalusite, cordierite and prismatic sillimanite in pelites and grossular garnet, idocrase and wollastonite in impure limestones. Between Crathie and Glen Gairn and on the south-east side of the Lochnagar Granite the hornfelses contain evidence of two thermal events.

The major intrusions were accompanied by widespread emplacement of dykes, sheets and veins ranging in composition from lamprophyre to aplitic microgranite, but dominated by microdiorite and felsite. There are at least two suites of microdiorites; the majority postdate the dioritic, granodioritic and early granitic phases, but do not cut the Lochnagar L2, Glen Gairn G2, Coilacriech and Ballater granites.

The district contains a number of sizeable faults all of which postdate the deformation of the Dalradian, but show equivocal relationships with the Newer Granites. For instance, the Glen Doll Fault, which brings Crinan Subgroup and Southern Highland Group rocks into contact, does not displace the outer margin of the Glen Doll Diorite, although the intrusion shows some evidence of brittle fracture. This fault is also truncated by the earliest phase of the Lochnagar Granite.

Uplift of the Dalradian, as evidenced by the retrograde metamorphism associated with the fourth phase of deformation, probably occurred in the Ordovician soon after the third deformation. Further buoyant uplift may have accompanied emplacement of the Newer Granites. Erosion of the Dalradian continued well into Devonian times and the presence of pisolitic silcrete of probable Devonian age on the north side of Glen Clova indicates that part at least of the present erosion surface was exposed at that time. The last significant pre-Quaternary geological event was the intrusion of quartz-dolerite dykes in Permo–Carboniferous times. Since none of these dykes apparently reached the then land surface it can be concluded that erosion continued into the Mesozoic.

In common with other parts of north-east Scotland, the Ballater district was subjected to deep subtropical weathering during the Miocene, when bedrock was commonly reduced to a coarse sand or gruss (Plate 2). Although much of the unconsolidated material was removed during the Quaternary glaciation patches extending to a depth of several metres have survived locally. During the late Devensian glaciation most if not all of the area was ice-covered and the pre-existing topography modified, most obviously by the creation of deep parabolic valleys. During glaciation large volumes of till and moraine were deposited some of which remains in Glen Muick and lower Glen Clova. However, much of this was reworked during the subsequent melting of the ice to form fluvioglacial sand and gravel spreads in the major river valleys. The melt-water also created a number of spectacular glacial overflow channels such as that occupied by the Burn of Vat [NO 42 99] (Crane, 1987). The landscape was further modified in the last major ice re-advance during the Loch Lomond stadial, with the development of corrie glaciers on the higher slopes. Since then depressions and river valleys have been partly infillled by alluvium and peat. Blanket peat also covers some of the upland areas, most notably to the south and east of Glen Muick.

Chapter 2 Applied geology

The tourist industry, based primarily on the impressive mountain scenery and on royal patronage of the upper Dee valley, is now the single most important employer in the Ballater district. However, little is done to market the area on the basis of its highly varied and readily accessible geology. There would seem to be considerable potential within the existing tourist framework for the creation of display boards, geology trails or even guided walks. There could also be a market for local rocks and minerals for decorative and ornamental purposes or even simply as souvenirs.

The key geological issues in the district are the rocks and minerals of past and present economic interest, and are summarised under the headings:

Construction minerals

Hard-rock aggregate

The district contains a wide range of rock types, many of which should be suitable for general purpose aggregate use and, although no physical test data is available for the area, some may even be capable of meeting the more exacting requirements for specialist end uses. In particular, some of the granitoid rocks should meet the specifications for low-shrinkage concrete aggregate. There are currently no working hard-rock quarries in the district and all roadmetal has to be imported. Since this involves road haulage over significant distances (a minimum of 20 km in the north, and in excess of 40 km for the Glen Clova area in the south) the demand for a local source could easily arise. There are a number of small disused roadside quarries which could readily be reopened; for instance in Ballater Granite [NO 3893 9645], and in Coilacriech Granite [NO 3055 9638]; [NO 3047 9584], in Glen Doll Diorite at [NO 288 758] and at [NO 3293 9650] where the dominant lithology is a porphyry dyke.

The numerous unmetalled hill tracks in the district are largely maintained from roadside borrow pits in sandy till or, more rarely, in decomposed granite or diorite for example at [NO 2699 9300] and [NO 2726 9256]. Barrow and Cunningham Craig (1912) noted that, due to its high soluble silica content, decomposed granite was the best material for binding unmetalled roads. Felsites, on the other hand do not suffer chemical weathering, but since they are very brittle and shatter easily to form small fragments they are also a valued local source of roadstone for hill tracks.

Dimension and polished stone

In the late 19th and early 20th centuries, agriculture and granite quarrying were the principal industries in the district (Barrow and Cunningham Craig, 1912). Although modest by comparison with other parts of north-east Scotland, the quarrying industry here nevertheless enjoyed the distinct advantage of minimal overburden thickness. Activity seems to have been concentrated in the Ballater Granite, with several quarries being worked near Cambus o’ May for example at [NO 398 987] and in the Lochnagar L2 Granite on the south-east side of Canup [NO 2407 9267] and at Inver [NO 225 931] just west of the district (Sheet 65W Braemar). Although some material was exported from the district, even as far as Aberdeen, the majority of the quarried stone was used locally. It is known that the granite for Balmoral Castle was wrought from the Canup quarry whereas Crathie Church is built of Inver stone.

There is no record of polished stone ever being produced in the district, but the wide jointing and crystalline nature of most granites and some of the calcsilicate-rocks should render them capable of producing attractive ornamental stone.

Sand and gravel

The Dee valley between Braemar and Ballater contains an estimated 14 million tonnes of sand and gravel (Peacock et al., 1977) much of which probably occurs within the district. The resource is spread over a large area and in consequence there is no single large deposit. The main concentrations are located in three areas on the south side of the Dee: immediately west of Ballater [NO 35 96], due south of Coilacriech Inn [NO 31 96] and [NO 32 96] and in the Feithan Laoigh valley south of Ripe Hill [NO 24 90]. Smaller deposits occur between Balnaut [NO 244 946] and Mains of Monaltrie [NO 243 953], to the west of Crathie, sporadically on the north side of the Dee between Crathie and Bridge of Gairn, in the lower reaches of the Gelder Burn [NO 24 93] and in lower Glen Muick, for example around Brochdhu [NO 352 932]. Sand and gravel is also present in Glen Clova to the south-east of Clova village. There are no working gravel pits at present in the district. However, there is evidence of extraction in the recent past at a number of sites, including [NO 353 960] just west of Ballater, [NO 3748 9629] 500 m north of Bridge of Girnock and 200 m south-east of Invergelder at [NO 242 935]. A further disused gravel pit is present on the north side of the A93, 400 m east of Crathie Church [NO 269 949].

The deposits encompass areas variously mapped in the primary survey as moraine, fluvioglacial gravel, kame and alluvium, but it seems likely that most, if not all, are glaciofluvial in origin. They occur principally in the form of winding mounds or flat-topped ridges, but also include terrace features, most notably that immediately west of Ballater which occupies at least 0.75 km2. Minimum thickness, as determined from several disused pits, is between 3 and 6 m. The resource consists largely of poorly sorted gravels with rounded to subrounded cobbles, up to 0.5 m diameter, of granite, psammite, quartzite and metabasic rocks. The gravels have a sandy matrix and layers and lenses of well-bedded sand are locally present.

Industrial minerals

Limestone

Limestone was formerly quarried for agricultural purposes at a number of locations in the Crathie and Ballater areas. The most extensive workings are in Blair Atholl Subgroup limestones, most notably the Crathie Limestone Quarry on the south face of Creag a Chlamhain [NO 2689 9549] where the face is up to 40 m wide and 50 m high, and the smaller quarry on the north side of Tom Buailteach [NO 2757 9385] where the worked face is 25 m long, but is less than 10 m high. Significantly smaller quarries were developed in Blair Atholl Limestone on the west side of Knock of Lawsie [NO 2594 9649], in Ballachulish Subgroup limestones [NO 248 961] on Creag Mhór, and on the south bank of the Dee by the Balmoral Castle Boat Pool [NO 2479 9548], and in Easdale Subgroup limestone close to the summit of Creag Phiobaidh [NO 3270 9472] (Plate 3) and 750 m south-east of Creagan Riabhach at [NO 375 985]. Most of the extraction appears to have been pre-20th century since none of the quarries was reported as working by Barrow and Cunningham Craig (1912). However, there is evidence that the Crathie Quarry was working at the time of the First World War (J Angus, personal communication, 1989), and the Tom Builteach quarry was not recorded by the primary survey and so presumably was developed after about 1900.

Baryte

Coarse-grained pink baryte is present in fault breccia forming two ribs in Allt Chernie [NO 3580 9030]. The mineral is present as a vein filling, in coarse tabular crystals up to 2 cm across, and has associated minor orange, finer grained celestite. In places the fault breccia is rotten and ochreous, and although no sulphide minerals remain, formerly they may have been present in small quantities. The total rock volume in the fault breccia here is approximately 5 cubic metres, of which no more than 5 per cent is baryte. The nature of the fault breccia and the north-east orientation of the faults are unusual for this area; no baryte has been found in the typical more siliceous fault breccia elsewhere. The area is poorly exposed and it is possible that the breccia extends beneath the drift.

Fluorite

Fluorite and calcite are the main gangue minerals at the Abergairn lead mine (see below). The fluorite is usually green, but purple, pale blue, white and rarely yellow varieties also occur. It ranges in form from coarsely crystalline to granular; it is intimately intergrown with calcite and may contain small crystals of galena and sphalerite. There is no indication that the mineral was ever worked, nor is there any record of reserves. Lawrie (1943) considered that an appreciable quantity of fluorite could be extracted from the dumps, but the intimate association of the mineral with calcite and country rock would necessitate some beneficiation process.

Lawrie (1943) also recorded two other fluorite occurrences in a cliff composed of Ballater Granite approximately 1.55 km east of Abergairn and 90 m north of the Monaltrie lead working (see below). The main one consists of a 30 cm-wide north-north-east-trending vertical quartz vein containing a 2.5 cm-wide central zone of green and purple fluorite with calcite and galena. Nearby white pyritous quartz with traces of greenish fluorite is present in a belt of shattered granite.

Andalusite

The three Al2 O5 polymorphs andalusite, sillimanite and kyanite transform into mullite at temperatures between 1350 and 1550ºC and are classed as high alumina refractories. The ability to form the mullite phase (which combines high strength at high temperatures with resistance to physical and chemical erosion) make these minerals valuable refractory raw materials. Of the three polymorphs andalusite is the most desirable for refractory manufacture since it has a volume increase on firing of only 3 to 6 per cent and hence does not require calcination prior to use.

Within the district aluminous pelite and semipelite units are most common in the Crinan Subgroup and the Southern Highland Group, but in contrast to much of north-east Scotland where andalusite is the prevailing polymorph, in the Ballater district the rocks are characterised by kyanite and sillimanite. However, as noted above, andalusite is developed within the thermal aureoles of the Newer Granites, particularly those surrounding dioritic bodies. The south-east side of the Lochnagar Granite is possibly the most promising area where the aureoles of the Glen Doll, Moulzie Burn and Juanjorge diorites overlap, although overgrowth of andalusite by sillimanite and K-feldspar occurs locally in this area. Other sites worthy of consideration include the north side of Glen Doll between Crags of Loch Esk and Cairn Lunkard, and the area immediately west of Ladder Burn around [NO 412 844].

Garnet

Garnet is used principally as an abrasive, although modest quantities find application in water filtration. Almandine and almandine–pyrope solid solution garnets give the best abrasive grades as they break down under pressure to give sharp, chisel-edged plates down to very fine grain sizes. They are used principally as abrasive powders and to manufacture coated and bonded abrasives for polishing wood, glass, rubber and plastics. Lower quality material may be used as an airblasting or hydroblasting media for metal cleaning.

Specifications for garnet generally focus upon the particle size range. Other considerations are garnet content (ideally as high as possible), free silica content (as low as possible), density, hardness and particle shape.

Garnet is widely distributed within the Dalradian rocks of the Ballater district and is most abundant in the Queen’s Hill Gneiss and in parts of the Southern Highland Group. However, within the Newer Granite aureoles the mineral is usually overgrown by cordierite. Hence the most promising areas are likely to be in the south-east part of the district.

Metalliferous minerals

The district contains only two occurrences, the Abergairn or Corrybeg lead mine and the Monaltrie workings, where metal is known to have been wrought in the past. However, there are indications of other mineralisation in the district, some of which may be economically significant. Much of the information stems from the results of the BGS Regional Geochemical Survey (now G-Base) which relies on the ability of stream sediment to detect even subtle changes in bedrock chemistry and to detect concealed mineralisation.

Vein mineralisation

Lead

Lead (Pb) was mined on a limited scale at Abergairn, 2 km north-west of Ballater at some time between 1820 and 1875. The old workings are situated 90 m north-east of Abergairn farm [NO 3550 9745] and were described in some detail by Dunham (1952). They are said to have extended to a depth of at least 10 m, although little remains on the surface other than a few shallow pits and some spoil heaps. The deposit lies within a north-north-west-trending, fault-controlled vein with several branches which can be traced for 100 to 110 m. The ore is reported to have been galena with a high proportion of silver, and sphalerite which occupied the central part of a calcite–fluorite gangue.

The host rock for the mineralised veins is pyritic schist of the Glen Girnock Calcareous Formation which is described in more detail below. The lead mineralisation is epigenetic and related to late-stage processes in the emplacement of either the Ballater or Coilacriech granites. Its close association with the Glen Girnock mineralised horizon is therefore largely coincident although it is possible that the strata-bound sulphides may have been a source of sulphur for the galena.

Lawrie (1943) located another vein approximately parallel to the main structure, but 135 m to the east-south-east. It consists mainly of white quartz with only traces of fluorite and galena.

The Monaltrie workings are located [NO 3704 9750] on the north side of the Pass of Ballater, and on the evidence of the primary survey appear to have been on a north-north-west-trending galena-fluorite-bearing quartz vein within the Ballater Granite. Lawrie (1943) states that the vein was worked for argentiferous galena, probably in the early 1840s, but at the time of his visit the workings were filled in and no trace of galena or fluorite was found.

On the south bank of the Dee [NO 3075 9615], at the junction between coarse and microgranite phases of the Coliacriech body there is a 2 to 3 m-wide, north-west- trending zone of intense quartz veining. The quartz is locally vuggy and may be coated with hematite. The vein is reputed to be rich in silver, a fact not confirmed by either the primary or this survey.

Tin

The evolved Coilacriech Granite shows substantial tin (Sn) concentrations in both stream sediments (217 ppm) and heavy mineral concentrates (3275 ppm) with associated Li, Be and Rb enrichment. The enrichment is considered to have resulted from extreme fractional crystallisation of a granite magma followed by greisenisation due to late-stage magmatic fluid activity (Plant et al., 1990; Tindle and Webb, 1989). Many of the high Sn values occur close to the greisen and vein-hosted W–Sn–Mo–Bi–Ag mineralisation at Gairnshiel (Webb et al., 1992) which is situated just north of the district (Sheet 75E). A second area of enrichment in stream sediments, 4 to 5 km west of Gairnshiel, may indicate further, but as yet unrecorded, mineralisation.

Disseminated mineralisation

Sulphides

A horizon of strata-bound sulphide mineralisation in the lower part of the Glen Girnock Calcareous Formation (Chacksfield et al. 1997) can be followed intermittently along strike for 7 to 8 km between Tullich Burn and Glen Girnock. The most extensive exposures lie to the north of Ballater, most notably on the southerly slopes of Creagan Riabhach where the horizon is around 60 m thick and can be traced along strike for at least 700 m. The mineralisation cannot be followed readily across the largely till-covered depression around the headwaters of the Loin Burn, but is visible in a trackside borrow pit [NO 366 983]. On the spur to the west, the horizon crops out at Craggans and at Abergairn [NO 3550 9745], where the exposure level has been enhanced by pitting in the area of the Abergairn lead mine. The outcrop width in the Craggans–Abergairn area is 70 to 100 m, which corresponds to a true thickness of 45 to 65 m.

The horizon reappears 3 km to the south-west, beyond the alluvial flats of the Dee valley, on the crags [NO 328 954] between Strathgirnock and Littlemill where it can be followed along strike for 400 m. The outcrop width, which is well constrained by good exposure in this section, is reduced to around 50 m. The only other exposure is in a 4 m-deep section on the east bank of the Girnock Burn [NO 3212 9417]. The occurrence is in an area of poor exposure, but it appears to lie within calc-schist which underlies massive amphibolite.

The mineralisation consists entirely of pyrite and pyrrhotite, no other sulphides being identified. The sulphides have a variety of forms including disseminated, thin, discontinuous cleavage-parallel trails and on joints. The principal hosts for the mineralisation are flaggy, bluish grey, calcsilicate-rocks, which include tremolite and actinolite schist. In the Girnock Burn occurrence, the greatest concentration of sulphides is in dark grey to black, possibly graphitic schist which forms a layer up to 1.8 m thick in the calcsilicate-rock succession. The sulphidic rocks are readily identified in the field by their distinctive weathered crust which is typically deep brown but also includes black and yellow. The sulphidic rocks show a range of magnetic susceptibilities from 0.38 to 7.17 SI, which presumably reflects variations in pyrrhotite content.

Sulphidic rocks occur elsewhere at this level in the formation, but not in the same concentration. For example, disseminated pyrrhotite occurs in actinolitic and tremolitic schist [NO 328 955] in the Strathgirnock–Littlemill section and in calcsilicate-rocks [NO 359 990] to the west of Creagan Rhiabhach. Pyrrhotite may also be present in rocks at the base of the formation as they can be distinguished as a magnetically anomalous zone.

The Glen Girnock sulphidic horizon, except for the presence of pyrrhotite, closely resembles the Pyritic Zone in the Ben Lawers Schist (Smith et al., 1984) both in character and lithostratigraphical position. The Pyritic Zone is laterally persistent, being traceable for over 150 km from Glen Shee to Loch Fyne but, with the exception of some very localised low-level copper enrichment, is largely devoid of base metals. This type of mineralisation, which strongly resembles the Norwegian Fahlbands (Gammon, 1966), probably resulted from sea-floor hydrothermal activity in relatively shallow water.

In the Crinan and Tayvallich subgroups fine-grained disseminated sulphides are common within some metabasic and calcsilicate-rocks, and may also be found within other lithologies. Pyrite is the commonest sulphide in the metasedimentary rocks, whereas the metabasic rocks contain both pyrite and pyrrhotite, which may form 1 to 2 per cent of the rock. The calcsilicate-rocks of the Water of Tanar Formation and to a lesser extent the Tarfside Psammite Formation, contain some 2 to 3 per cent disseminated sulphides, again mainly pyrite and pyrrhotite, although some chalcopyrite and bornite was also identified. In all cases, there is minor remobilisation of sulphides into late fractures and onto joint surfaces, but no large vein-hosted remobilised deposits have been located.

Copper

Coincident anomalous concentrations of copper (Cu) (about 200 ppm) and lead (about 150 ppm) were recorded in stream sediments overlying Crinan and Tayvallich subgroup rocks to the east of Lochnagar (British Geological Survey, 1991). Similar values of copper are accompanied by abundant pyrite in heavy mineral concentrates over the same rocks between Lochnagar and Glen Doll. The enrichment, which is also evident in the contaminated dioritic rocks of Glen Doll (Jarvis, 1987) may indicate the presence of minor strata-bound sulphide mineralisation, although this was not confirmed in the rocks.

Zinc

High values of zinc (Zn) 600 ppm) occur in stream sediments over the Queen’s Hill Gneiss Formation to the east of Lochnagar. The Glen Doll dioritic rocks are also characterised by enhanced levels of zinc (<160 ppm) which, according to Jarvis (1987), result from the assimilation of locally derived Dalradian zinc-rich metasedimentary xenoliths. Schumacher (1985) and Goodman (1993) recorded staurolite with up to 2.4 per cent ZnO in Glen Doll and Glen Muick. The latter author noted that the rocks include pyrite (with 2% Co) and minor chalcopyrite instead of the usual oxide suite, and concluded that the unusual composition of the rock might reflect minor exhalative mineralisation.

Gold

Coats et al., (1993) recorded gold (Au) in rocks from the Fleurs Fault Zone (Chapter 16) where it is exposed in the Burn of Fleurs [NO 3899 7304]. Values of up to 7 ppm gold were obtained from the clay with quartz-rich stringers which forms the main fault gouge. The marginal limonitic material was also enriched, with up to 1.6 ppm gold.

The same authors also found minor gold anomalies in panned concentrates collected from two streams on the south side of Glen Doll, the Burn of Kilbo and the Fee Burn, which drain Southern Highland Group gneisses and the southern part of the Glen Doll Diorite (Coats et al., 1993). One such anomaly was traced to a 40 mm-wide quartz vein cutting gneissose semipelite [NO 2705 7491]. The vein contains sporadic pyrite, chlorite and minor pink feldspar with up to 1.7 ppm gold.

On the evidence that gold had historically been panned in Allt Deas [NO 38 88] and the central Water of Tanar [NO 39 89], a suite of panned concentrates was collected from the Water of Tanar catchment area. No visible gold was found, but analysis of the panned concentrates detected significant amounts of gold, up to 100 times background in some samples. The sample showing greatest enrichment was collected just downstream from an outcrop of vuggy fault breccia, with two further samples from the same location both giving anomalously high gold values. Farther downstream, gold was still detected, although in lower concentration; however a tributary stream nearby, draining only the granite, showed no appreciable gold. Likewise samples collected from the streams running over the Tarfside Formation and the Water of Tanar Limestone showed no enrichment in gold. The results indicate that gold may be associated with the north–south faulting, although these preliminary results should be interpreted with care, owing to the ‘nugget effect’ prevalent in gold analyses. The results do, however, suggest that the area is prospective for gold.

Chapter 3 Concealed geology

Geophysical data

Gravity data

The Bouguer gravity anomaly map of the region (Figure 3) is strongly influenced by the relatively low density granitic intrusions. Bouguer gravity anomalies over the main granites have minima of less than −53 mGal over Coilacriech where it crosses the River Dee, through about −51 mGal over Ballater, and about −50 mGal over Lochnagar, just west of Ripe Hill and over the western part of Mount Battock. Maximum anomalies of about −10 mGal occur in the south around Glen Clova, although a ridge in the gravity data along the east side of the Lochnagar Granite can be traced from the Glen Doll Diorite [NO 280 770], through the Cul nan Gad Diorite [NO 332 874] to just west of the Coyles of Muick basic–ultramafic body.

Lineations picked from images of regional gravity and magnetic data (Figure 3) suggest a north-north-east-trending structure extends for over 20 km from Juanjorge to the south-east side of the Ballater Granite. This structure may have in part controlled the south-east margins of the Lochnagar and Ballater granites and for part of its length correlates with the Glen Doll Fault. Clear inflexions in the contours trend south-eastwards from the Glen Doll Diorite, and indicate the positions of two north-west-trending regional structures. The more southerly runs along the southern flank of Glen Clova and along Glen Doll; the second parallels the valley floor in Glen Clova, crosses the Glen Doll Diorite and thereafter follows the south-west margin of the Lochnagar Granite. To the north-east of this, subtle gradient features in the Bouguer gravity anomaly map indicate a series of north-west-trending structures crossing the Dalradian sequence. These can be related to major changes in strike of the Southern Highland Group rocks (Figure 3) and extensions of the structures controlling the north-east and south-west margins of the Lochnagar Granite. There is only one regional gravity station on the Coyles of Muick basic–ultramafic mass which explains why this body is not particularly evident in (Figure 3).

Aeromagnetic data

Aeromagnetic data over the region were collected at 305 m above terrain in 1963, along east–west flight lines spaced approximately 2 km apart. This regional data has been digitised and the main features of the total field anomaly for the Ballater region are shown in (Figure 3) superimposed on the gravity data. There is a prominent 350 nT anomaly over the Glen Doll Diorite (1 on (Figure 3)), but the most significant anomaly (500 nT) is over the contact between the Juanjorge Diorite and the Lochnagar L1 Granite (II). An annular aeromagnetic anomaly of about 100 nT which coincides approximately with the outcrop of L1 on the east side of Lochnagar (III) is resolved by the detailed ground magnetic data into a number of separate features relating to the various marginal diorite bodies. The magnetic data suggests that these diorites may be continuous at depth. A single flight line crosses the Coyles of Muick body and indicates a prominent anomaly of 150 nT (IV). Another significant aeromagnetic anomaly, with values above 200 nT, occurs to the south-west of the Mount Battock Granite, just east of the contact between the Tarfside Psammite and the Longshank Gneiss formations (V). Again, this is resolved by the ground data into two clear units of magnetite-bearing migmatitic gneiss within the lower part of the Longshank Gneiss. A similar anomaly just north of Loch Wharral (VI) is considered to be due to similar source material. A minor positive anomaly occurs over Cairn Trench and Gibs Knowe, and there is a further local maximum to the east of the district [NO 344 772].

Ground magnetic data

The results of detailed ground magnetic traverses across the central part of the district, where the solid rocks are, in general, poorly exposed have been described by Goodman et al. (1990) and Tadesse (1991). The ground magnetic data, observed at intervals of about 20 m along about 900 km of traverses 200 m apart, cover most of the metsedimentary rocks north of the Glen Doll Diorite and south of Ballater. These data have been replotted relative to a datum of 49 500 nT and are shown in (Figure 4). Most of the significant aeromagnetic anomalies are more clearly mapped by the ground data apart from the Loch Wharral aeromagnetic anomaly, [VI] which is not within the area of the ground surveys.

A strong linear ground magnetic anomaly (with local values of about 1400 nT) is associated with the sulphidic horizon in the lower part of the of the Glen Girnock Calcareous Formation, described in Chapter 2. The other major zone of high amplitude and high frequency ground magnetic anomaly occurs in the Longshank Gneiss Formation to the south-west of the Mount Battock Granite. Between the granite and the Water of Mark, the Tarfside Psammite–Longshank Gneiss interformational boundary is clearly delineated; farther south-west the boundary is less clear, but the form of the anomalies suggests that the Longshank Gneiss underlies the Tarfside Psammite. North of the Water of Lee [NO 396 803], two clear magnetic horizons with a general north-north-east strike are detectable in the Longshank Gneiss Formation. This magnetic pattern is not seen on Hunt Hill [NO 380 805] and the truncation of subparallel anomalies can be interpreted as indicating a major fault trending north-west along the Water of Lee.

Many of the Argyll Group metabasic bodies have enhanced magnetic susceptibilities and give rise to significant ground magnetic anomalies of at least 1000 nT (Table 2). East of the Coyles of Muick [NO 328 910] the magnetic metabasic units are arranged in two main zones, Allt Darrarie [NO 320 828]–Allt Downie–Cairn Leuchan [NO 830 911] and Linn of Muick [NO 332 895] to Craig Vallich [NO 377 926]. A strong (800 nT) anomaly over the metabasic body situated 2 to 2.5 km south-east of Loch Muick [NO 290 830] has been interpreted by Goodman et al. (1990) as indicating a large-scale F2 fold with a lower limb dipping to the north-west. The unit is assumed to have an induced magnetisation of about 1.8 A/m, and extends to about 1 km depth, with a mean susceptibility of about 45 x 10−3 SI. Other prominent anomalies are associated with the sheared metadiorite bodies in Glen Mark [NO 365 853], and east and west of Am Mullach [NO 375 903]. The Glen Mark body is modelled as a wedge-shaped structure dipping to the north-west with a mean susceptibility of 20 to 45 x 10−3 SI.

The amphibolitic unit in the lower part of the Tarfside Psammite Formation is magnetic in part and as such forms a useful marker which shows significant local dislocation caused by late faulting. This unit can be traced magnetically north-eastwards from Muckle Cairn [NO 353 825], through Hare Cairn [NO 377 880] to Knockie Branar [NO 401 928], but only extends south-westwards for about 1 km. However, in the upper Capel Burn [NO 330 790] north-east of the Glen Doll Diorite, minor magnetic anomalies are associated with metabasic rocks at the junction between gneisses of the Queen’s Hill and Cald Burn formations. Correlation of the Capel Burn and Muckle Cairn amphibolitic rocks would suggest that the Cald Burn Gneiss Member is the lateral equivalent of the Glen Tanar Quartzite Member.

The ground magnetic data indicate that the Moulzie Burn Diorite is separated from the Juanjorge Diorite by a screen of metasedimentary rocks, and that it is probably continuous at depth with the Glen Doll Diorite. The ground anomalies over the Allt Darrarie and Cul nan Gad diorites could be interpreted as indicating a near continuous diorite–tonalite body around the east side of the Lochnagar Granite.

Geophysical interpretation

Major lineaments identified from the regional gravity and aeromagnetic data together with fault structures interpreted from the detailed ground magnetic data are shown on (Figure 3).

(Figure 5) shows a model profile based on the regional geophysical data along most of the cross-section on the geological map, extended by approximately 16 km at either end to include the effects of the Glen Gairn and Mount Battock granites. The significant regional features are the marked increase in the regional aeromagnetic anomaly to the east of the Coyles of Muick, the broad zone of low gravity anomaly over the Cairngorm and Glen Gairn granites and the gravity minimum close to the south-west side of the Mount Battock Granite. The central part of the cross-section, east of the Coyles of Muick, is modelled as about 2 to 3 km of Dalradian strata above granite at depth. There must be significant off-section effects from the granites to the north of this line so that the model is considered as only approximate. The gravity profile shows a marked change close to the sheared metadiorite just west of Am Mullach, and it seems likely that most of the ground east of Am Mullach is underlain at shallow depth by Mount Battock Granite. Tadesse (1991) modelled ground magnetic data across the Coyles of Muick mass as a series of magnetic units dipping generally to the south-east, with a magnetisation of up to 2.5 A/m and an implied mean volume susceptibility of about 60 x 10−3 SI. A reasonable fit was obtained assuming a depth of about 600 m to the base of the ultramafic rocks; this was considered a likely depth on structural grounds.

(Figure 6) shows an interpretation of the ground magnetic data along the central part of the section. A simple model of the main features suggests an eastward dip for the formations west of the Coyles of Muick and a steep or vertical dip for the mafic rocks. A significant but restricted ground anomaly close to Toldhu on the section line [NO 346 912] might be due to mafic rocks at depth rather than metabasic rocks nearer surface. The prominent ground anomalies associated with the metabasic rocks of Craig Vallich [NO 377 926] and Cairn Leuchan [NO 380 911] do not extend south to the section line. Despite a fairly consistent steep regional dip to the north-west at outcrop the ground magnetic anomalies across the metabasites of Craig Vallich suggest a dip to the south-east, as do many of the anomalies on the line of cross-section. Goodman et al. (1990) considered that the anomaly pattern reflected a F2 fold with an upper limb dipping gently to the east, and a steep westerly dipping lower limb.

In the south-west Grampian Highlands the sequence of intrusion and mode of emplacement of the major post-tectonic granitoids is controlled by the intersection of late Caledonian south-west–north-east transpressional shears and early pre-Caledonian north-west features and lineaments (Jacques and Reavy, 1994). A similar pattern of structural control, with emplacement by localised ballooning also probably occurred in parts of north-east Scotland including the Ballater district. For example, the Lochnagar Granite is bounded by north-west-trending gravity lineaments, comparable with those in the south-west Highlands, and on the south-east side by a prominent Caledonian shear zone. The geophysical evidence suggests that Lochnagar, like the intrusions of the south-west, is a steep-sided mass extending to a depth of about 10 km. In contrast, the western part of the Mount Battock Granite might have been controlled by transpression on the north-north-east-trending shear zone and have a diapiric form in the upper crust. Whether granite extends at depth between Lochnagar and Mount Battock is a considerable problem to resolve given the total volume of granite in the east Grampians and the considerable gravitational effect of this. The model shown in the section (Figure 5) suggests that the Mount Battock Granite and possibly buried parts of the Ballater Granite might extend west to the Coyles of Muick Shear Zone, and that there may be a thin screen of country rock separating this from Lochnagar.

Borehole data

Detailed geochemical, geophysical and petrological studies on a 298 m-deep borehole [NO 401 986] in the Ballater Granite have been described by Lee (1984) and Webb and Brown (1984). The borehole was sunk as part of a programme to investigate the Hot Dry Rock geothermal potential of selected Caledonian granites.

The upper 30 to 40 m of core shows mild alteration, possibly due to deep weathering, and has a mean saturated density of about 2.5 Mg m−3 compared with about 2.6 Mg m−3 elsewhere in the core. Below this the granite is largely fresh, although slightly altered and jointed sections were noted at 163 to 181 m and 225 to 244 m, and below 280 m the core is severely jointed and hematised. The granite core displays an increase in density and decrease in heat production with depth which accords with the geochemical evidence that the granite is slightly less evolved towards the base of the hole. In terms of heat production the core gave a mean of 6.9 µW m−3.

Chapter 4 Dalradian—stratigraphy and depositional environment

The succession Recognised within the Dalradian rocks of the Ballater district (Figure 7) is lithostratigraphical, although there are some chronostratigraphical markers such as the Boulder Bed, the Meall Dubh and Balnacraig metabasites, the Green Beds of the Rottal Schist Formation and the pyritic schists in the Glen Girnock Calcareous Formation. With the virtual absence of coherent sedimentary structures the stratigraphical order in the district is based on comparisons with Dalradian rocks elsewhere in Scotland. In particular, the Appin Group and lower Argyll Group stratigraphy is based largely on analogies with Upton’s (1986) succession in the adjacent Braemar district (Sheet 65W), whereas the remainder relies on relationships established in the Pitlochry district (Sheet 55W) (see Harris et al., 1994).

It will be evident from (Figure 7) that, whereas the Ballater district stratigraphy can be readily correlated with the established Dalradian succession at group and even subgroup level, for a number of reasons individual formations cannot always be readily matched. This is particularly so for the Appin Group formations. In the case of the Ballachulish Subgroup, this is due to an incomplete succession resulting from fragmentation of the outcrop by major granitoid intrusions. For example in the Braemar area, Upton (1986) recognised four lithostratigraphical units between his An Socach–Cairnwell Quartzite and the Gleann Beag Dark Schist and Limestone Formation, whereas around Crathie the stratigraphically equivalent succession of schistose semipelite with interbedded calcsilicate-rock, limestone, pelite and minor quartzite is termed the Birchwood Semipelite Formation. However, within the formation, three lithologically distinct limestone members have been identified, two of which, the Glen Baddoch and Gleann Beag, are equated with limestone units in Upton’s succession. Also, the overall distribution of the large metasedimentary xenoliths within the granitoids suggests that there has been little or no relative movement between them, so locally a ghost stratigraphy is preserved, particularly within rocks of the Blair Atholl Subgroup.

Another reason for the absence of bed-by-bed correlations is lateral facies changes, a well-known feature of the Argyll Group, particularly within the Easdale Subgroup. This is evident in the absence of graphitic schist low in the subgroup (compare with Ben Eagach Schist in Perthshire), that part of the succession in the Ballater district being represented by psammite and semipelite of the Creag nam Ban Formation. Further lateral facies variation is evident in the along-strike transition within the Tarfside Psammite Formation from the Glen Tanar Quartzite Member to the Cald Burn Gneiss Member.

The Dalradian has long been recognised as a thick sequence of metamorphosed marine sedimentary and volcanic rocks (Harris et al., 1978; Anderton, 1985). Whereas the Appin and Argyll groups accumulated in an ensialic basin those of the Southern Highland Group are thought to have prograded onto oceanic crust (Soper and England, 1985). The protolith of the Appin Group comprised a mudstone–limestone–sandstone sequence which was deposited in a quiescent, gently subsiding, open, stable shelf environment such that individual formations, for example the Cairnwell Quartzite Formation, can be followed for considerable distances. Sedimentary structures are confined to a few examples of graded bedding in the quartzite and some possible load structures in Blair Atholl Subgroup rocks.

The Boulder Bed is considered to have originated as a tillite, which explains its somewhat sporadic occurrence. It is traditionally taken to mark the base of the Argyll Group, but because of its absence over much of the Ballater district the lowermost unit of the group is generally the Creag Leacach Quartzite Formation. Deposition of the Argyll Group was characterised by an upward increase in tectonic instability with deposition in faulted second and third order basins (Coats et al., 1984) and intermittent volcanism. Even the Creag Leacach Quartzite Formation, which is one of the laterally most persistent units in the group, is absent for a short stretch on the north side of the Dee [NO 28 91]. This occurs on the limb of a major F2 fold and could be the result of tectonic thinning but, coupled with the predominance of psammite and semipelite and the absence of graphitic schist in the overlying Creag nam Ban Formation, it is thought to reflect the presence of a topographical high and hence the lack of sedimentation in the lower part of the Argyll Group. The Glen Girnock Calcareous Formation is essentially clastic in origin, but the fine-grained sedimentary rocks and some of the limestones are considered to have a chemical component. The formation also includes a considerable thickness of hornblendic rocks some of which may have had an extrusive origin, providing the earliest evidence in the district that sedimentation was periodically interrupted by volcanism. More obvious and more widespread evidence of contemporaneous magmatism is provided by the overlying Meall Dubh Metabasites Formation, the Balnacraig Metabasites Member within the Queen’s Hill Gneiss Formation and the Green Beds of the Southern Highland Group.

The base of the Crinan Subgroup (Queen’s Hill Gneiss Formation) marks an important change in the sedimentary history of the Dalradian, with the restricted basins giving way once more to more open conditions, although now with high-energy turbiditic-type deposition, probably in a more distal environment. Such conditions characterised deposition throughout much of upper Argyll Group and Southern Highland Group. This change in depositional environment is quite rapid, but as the original nature of the junction is masked by strong shearing it is not clear whether it represents a true sedimentary passage.

The Crinan Subgroup is composed largely of metagreywacke, but it encompasses a south-west passage from a mixed semipelite–pelite–psammite succession to a more monotonous pelitic and semipelitic succession. This passage most probably represents a transition from a shelf edge to a more distal environment, although the presence of psammite farther to the south may indicate that this was not a major basin margin. This change in lithology is quite subtle and takes place in the area between Round Hill of Mark [NO 33 82] and Allt Darrarie [NO 32 83]. It occurs across the same north-west-trending lineament defined by more obvious facies variation higher up the succession (Figure 8). The effect of the lineament is most pronounced in the Tayvallich Subgroup, with the rapid transition within the Tarfside Psammite Formation from quartzite and psammite in the north-east (Glen Tanar Quartzite Member) to semipelite and psammite in the south-west (Cald Burn Gneiss Member), and in the south-westerly wedging out of the Water of Tanar Limestone Formation. This may be variously interpreted as marking a change from distal shelf to shelf edge or simply from basin high to basin deep. This lineament may represent a sub-basin margin that was active throughout the deposition of Crinan and Tayvallich Subgroup rocks. It also appears to have influenced the distribution of metabasic sheets in the Tarfside Psammite Formation for although these rocks are common in the Glen Tanar Member they are not present in the Cald Burn Member. However, there is no corresponding control on distribution of metabasic sheets in Crinan Subgroup and Southern Highland Group rocks either side of the lineament.

The line was still active during deposition of the Longshank Gneiss Formation though the effect was much more subdued than during the deposition of the Tayvallich Subgroup, possibly because deposition occurred under a greater depth of water or possibly because there was less clastic input to the area. Depositional conditions were analogous to those of the Crinan Subgroup, with a subtle lateral facies variation across the lineament. Mixed semipelite and pelite with common psammite bands occur in the north and east (for example around Drumhilt [NO 35 80]) passing into a more monotonous sequence with only rare psammite in the south and west, (for example on Cathelle Houses [NO 31 76]). The volume of interbedded psammite increases again to the south-west, mimicking the variations noted in the Crinan Subgroup. The termination of the Green Beds to the south of Cairn Trench [NO 38 73], near the south-easterly extension of the lineament, and their absence to the north-east indicates that it continued to influence deposition throughout much of Southern Highland Group.

The precursors of the rocks that make up the bulk of the Longshank Gneiss Formation are considered to have been distal turbidites. However, around Driesh [NO 27 73] the lower part of the formation is dominated by psammite (Driesh Psammite Member) which is interpreted as a local, more proximal, sandy facies. The member is adjacent to the Glen Doll Fault, and there are other significant changes in stratigraphy within the Tayvallich Subgroup and Southern Highland Group across the fault which suggest that it, or its precursor structure, acted as a sub-basin bounding structure at the time of sedimentation.

The Rottal Schist Formation is dominated by gritty psammite and marks a change to a more proximal depositional environment. In contrast to the Longshank Gneiss Formation, it contains metabasic rocks, including volcaniclastic metasedimentary rocks and metadolerite sheets which become increasingly dominant to the west.

Chapter 5 Dalradian — Appin Group

The Appin Group succession in the Ballater district comprises rocks of the Blair Atholl and Ballachulish subgroups (Figure 7). They are the most restricted of the Dalradian rocks in outcrop area being confined to the ground bounded by the rivers Dee and Gairn (Figure 9). They occur predominantly as screens and roof pendants in the Newer Granites and to a lesser extent within the younger basic suite, so the outcrop of individual formations cannot be traced with certainty for any great distance. However, the overall disposition of the xenoliths suggests that there has been little or no relative movement between them; locally a ghost stratigraphy is preserved, particularly within rocks of the Blair Atholl Subgroup. Because of proximity to intrusive bodies the Appin Group rocks are invariably hornfelsed.

Ballachulish Subgroup

Cairnwell Quartzite Formation (QQB)

This formation is equated with the An Socach–Cairnwell Quartzite which crops out extensively in the Braemar district to the west. However, in the Ballater district it is restricted to a small lobe within Abergeldie Complex diorite, which extends eastwards from the western edge of the map for less than a kilometre. The quartzite is well exposed on the eastern slopes of Carn Moine an Tighearn [NO 33 96] to [NO 23 97], but to the north of this the position of the outcrop is largely traced from surface debris or float. Typically the quartzite is clean and feldspathic, glassy and pale brown where fresh, but white to pink on weathered surfaces, with some colour banding. Its feldspathic nature is most obvious on weathered faces; the white-weathering feldspar grains, which may show evidence of flattening, are disseminated throughout the rock and are, in places, concentrated in layers. Bedding is generally defined by dark heavy mineral bands, and more rarely by feldspar-rich layers or micaceous partings. Graded bedding was recorded in the quartzite at a number of localities, and in one instance [NO 2313 9666] is accompanied by centimetre-scale cross-bedding.

Birchwood Semipelite Formation (bSCB)

This formation is thought to underlie most of the topographical low that extends north-westwards from Balnaut [NO 244 945], although in practice much of the outcrop is obscured by head or other superficial deposits, and exposure is confined largely to the tributaries of the Coldrach Burn. The most extensive exposures are on the ridge running north-westwards from Creag Mhór as far as the Crathie Burn. To the north of this, the position of the formation cannot be established with any certainty, although isolated patches of semipelite, psammite and white limestone have been found in the Duchrie Burn [NO 2336 9894]. On the north-eastern slopes of Craig Nordie [NO 23 94] numerous screens of semipelite and calcsilicate-rock are present within the diorite, whereas the only exposures recorded south of the Dee are two patches of limestone within diorite [NO 248 946] and [NO 254 943]. Because of their intimate association with major intrusions, these rocks everywhere bear the hallmarks of strong thermal metamorphism.

The Birchwood Semipelite Formation consists largely of semipelite and pelite, but includes three lithologically distinct limestones, the Baddoch Burn, Gleann Beag and Creag Mhór limestone members. Also, the succession is dominated locally by quartzite. For example west of Creag Mhór between [NO 236 966] and [NO 240 972] the formation consists largely of white quartzite which, with its feldspathic nature and heavy mineral bands, is not unlike the Cairnwell Quartzite Formation. However, the presence of interbedded flaggy psammite and semipelite and the position relative to the limestone indicate that it is part of the Birchwood Semipelite Formation. The semipelite and pelite are typically dark bluish grey and massive, but in parts there is a spaced cleavage or, more commonly, irregularly spaced parallel fractures. They are generally coarse grained, and in places almost gneissose with thin pale grey psammite layers. The semipelitic rocks comprise cordierite, red-brown biotite, garnet, quartz and plagioclase. Quartz in the psammitic layers is accompanied by finer grained granular feldspar including plagioclase and K-feldspar. The more pelitic rocks (for example (S77669)) have an essentially similar mineralogy but lack K-feldspar, and have the additional presence of sillimanite and possible andalusite.

On the lower eastern slopes of Craig Nordie, and in the patchy exposures to the north-east numerous metasedimentary screens occur within granodiorite and diorite. These show a much greater lithological diversity than those on Creag Mhór, and consist of semipelite, calcsilicate-rock, micaceous psammite and pelite. Evidence from the eastern edge of the Braemar district indicates such features are characteristic of the sequence above the Gleann Beag Ribbed Limestone. The most extensive exposure is a 70 x 80 m screen centred on [NO 2307 9524]. It comprises, at its lowest point, a unit of calcsilicate-rock about 10 m thick with millimetre- to centimetre-scale light and dark grey colour banding that is locally discontinuous and results from alternating layers of plagioclase + quartz + diopside with those of hornblende + biotite. This rock is succeeded by micaceous psammite which, near the top of the screen, gives way to further calcsilicate-rocks and rusty weathering dark grey to black schist marking the base of the Blair Atholl Subgroup. A more homogeneous diopside + plagioclase rock occurs at [NO 2372 9536] close to a small granite boss; it has a mottled texture due to the local development of red-brown biotite and contains small rounded grains of possible corundum.

Baddoch Burn Limestone Member (LBB)

In the Braemar district (Sheet 65W) the An Socach–Cairnwell Quartzite is succeeded by the Glen Baddoch Dolomite (Figure 7). The only known exposures of this unit in the Ballater district are in two 15 to 20 m diameter, shallow, grassy depressions [NO 237 971], which resemble sink holes. The member consists of a coarse to very coarse crystalline white limestone, characteristically pure and carrying only minor amounts of quartz, feldspar, diopside and magnetite. The limestone gives off a fetid smell when struck with a hammer. The rock has a faint bedding, defined by iron oxide-rich layers that dip to the north-west at a low angle. No contact with the enclosing diorite is seen but a small exposure of pink equigranular granite is present at the north-west corner of the more westerly of the two depressions.

Gleann Beag Limestone Member (bLB)

This member consists of alternating soft carbonate and hard calc-silicate layers which give rise to a prominent ribbing on weathered surfaces. It is lithologically similar to the Gleann Beag Ribbed Limestone in the Braemar district (Upton, 1986). Its position in the succession of the Ballater district, particularly with regard to the Cairnwell Quartzite, was established with some certainty on Creag Beag [NO 223 947], which lies approximately 1 km to the west of the district.

In the Ballater district the member is best exposed on the south bank of the River Dee, adjacent to the Boat Pool [NO 248 946] in a near-continuous 70 m-long cross-strike section and an adjacent quarry face 6 to 7 m high. At this locality the unit dips consistently between 20 and 25° to the south and has an estimated outcrop thickness of 20 to 25 m. It consists essentially of a white to pale grey limestone with some mid-grey layers. The limestone is relatively pure with only minor amounts of diopside, sphene, plagioclase, quartz, epidote and skeletal garnet. Calcsilicate layers are more common in the lower part of the member; they are locally flinty and may be studded with idocrase porphyroblasts up to 10 mm long. The Boat Pool section also includes amphibolite units, both concordant and discordant, up to 0.6 m thick, which in places are disrupted and boudinaged and have infolded limestone. The uppermost bed seen is a mid-grey psammite flecked with biotite and with conspicuous iron staining on its weathered surfaces. The only other noteworthy exposure of the member is provided by a small disused quarry [NO 2479 9607] on the southern slopes of Creag Mhór, where a 10 m-wide face of pale grey and white crystalline limestone is preserved between microgranodiorite on the west and paler microgranite on the east. Near the top of the exposure the rock is noticeably ribbed with thin mica schist and amphibolite layers.

Creag Mhór Limestone Member (LM)

This member lies within the upper part of the formation and is confined to just two localities. At the first, it forms a small spur near the eastern end of the small disused granite quarry [NO 2448 9576] on Parliament Knowe, where it comprises an essentially flat-lying, white and pale grey colour-banded crystalline carbonate rock within which irregularly distributed flecks of biotite are present. The rock is in part brecciated and highly silicified, and is probably separated by a fault from the pelitic to semipelitic hornfelses which occur immediately to the east and 20 m upslope. At the second locality south of the Dee, it forms a low ridge running south-west from Prince Arthur’s Cairn [NO 2538 9529]. The exposure, over 100 m long and 30 to 40 m wide, comprises fine-grained, pale to mid-grey speckled crystalline limestone with poorly defined bedding dipping at about 30° to the north-east.

Blair Atholl Subgroup

Blair Atholl Subgroup rocks can be traced from the western edge of the district just north of Inver to Glen Gairn [NO 235 942] to [NJ 32 00]. The lithostratigraphy of the subgroup (Figure 7) is largely derived from the type section on Creag a’ Chlamhain [NO 26 95] and from exposures on Sròn Dubh [NO 29 96] where reasonably coherent sections are preserved, although the rocks are intruded by amphibolite, diorite and granite. No contact with Ballachulish Subgroup rocks is visible.

Crathie Schist and Limestone Formation (SPBA)

The presence of rusty weathering, black and, in places graphitic schist units and thick beds of relatively pure limestone are the features which serve to distinguish the rocks of this formation which also includes lesser volumes of semipelite, pelite, psammite and calcsilicate-rock. Although the outcrop is extensively interrupted by intrusions the following generalised lithostratigraphy has been established:

The lowest exposed sequence, of pelite and semipelite crops out on the south-western slopes of Creag a’ Chlamhain and consists principally of dark bluish grey, massive, cordierite-rich rocks in which a faint cleavage is preserved only locally. Two specimens ((S92671) and (S92672)) of semipelite [NO 2655 9526] contain andalusite and coarse prismatic sillimanite (visible in hand specimen), although cordierite and plagioclase are largely pseudomorphed by sericitic white mica, chlorite and pinite. Weathering of the abundantly disseminated iron sulphide in places imparts a rusty crust to these rocks. Paler grey, more iron-poor bands, possibly reflecting primary bedding, and some psammitic units are also present. These rocks may be traced to the cliff-lined south face of Creag a’ Chlamhain where their bedding planes, which dip to the north-east at gentle angles, effectively define the roof of the underlying quartz-diorite.

On Creag a’ Chlamhain, the lowest sequence is overlain by noticeably phyllitic pelite which, although not definitely graphitic, is dark grey to black in colour. On Sròn Dubh, the succession is broadly comparable with that described above, but the dark schist is quite black and more obviously graphitic. In thin section (S92842) the rock shows evidence of strong retrogression with the plagioclase and andalusite being entirely replaced by sericite. The essential mineralogy is quartz, biotite, cordierite, K-feldspar, sillimanite and late muscovite. Bedding is seldom discernible in these rocks, the principle fabric being a widely spaced irregular cleavage.

The dark schist succession includes thick units of laminated micaceous psammite comparable with that forming the overlying Lawsie Psammite Formation. Also at one locality [NO 2675 9538], near the base of the cliff on Creag a’ Chlamhain the dark schist includes a 200 mm-thick unit of quartzose psammite into which the overlying micaceous pelite has slumped. The schists are separated from the main limestone unit by a further thickness of semipelite which includes layers of cordierite–biotite rock (S82360). These layers have a distinctive spotted texture which results from biotite and cordierite containing localised concentrations of finely divided opaque minerals (possibly graphite).

The main limestone can be followed almost continuously for 300 m north-westwards from Crathie Limestone Quarry [NO 269 955], and thence intermittently for a farther 200 m through more poorly exposed ground. Continuing in the same direction across the largely drift-covered valley of the Rimarsin Burn, a further 400 m of near continuous strike exposure is present on the southern slopes of Knock of Lawsie between [NO 261 962] and [NO 259 966]. The outcrop width ranges from 30 to over 100 m, which on the well exposed part of Creag a’ Chlamhain corresponds to an estimated true thickness of 20 to 35 m. The main limestone also crops out on the northern summit of Sròn Dubh [NO 2933 9671], where it is between 21 and 22 m thick.

The most comprehensive section through the limestone occurs in Crathie Quarry where for the most part the unit dips uniformly to the east at 30 to 40°. Local variations in orientation caused by open F3 and F4 folds are evident within the quarry and close to the eastern wall the dip steepens to between 75 and 80°. The base of the limestone is seen at the south-west corner of the quarry where it overlies dark semipelitic hornfels, but the eastern (hanging) wall consists largely of feldspar-phyric amphibolite. The amphibolite locally transgresses the bedding in the limestone and is affected by F3 folds. The limestone in the quarry is of very uniform appearance with only rare interbeds of dark semipelite. It is composed of a well-bedded, pale grey crystalline rock with millimetre- to centimetre-scale mid to dark grey colour banding. In the field it is readily distinguished from the other limestones by its pale-grey weathering as opposed to the more usual tan and black colouring.

However, it is an impure limestone and in thin section (S92669) is seen to contain abundant plagioclase, tremolite, diopside, quartz, sphene, opaque minerals and possible forsterite. Heddle (1901) also recorded wollastonite, idocrase and fluorite at Crathie Limestone Quarry. The rock shows little variation along strike to the north-west; a further specimen (S82336) collected from a small track-side quarry on the west side of Knock of Lawsie [NO 2594 9649] contained no tremolite, but had in addition epidote, serpentine and magnesian chlorite. In this area pale and dark grey banded calcsilicate-rock and hornblende schist were recorded at or close to the base of the limestone. The main limestone shows more lithological variation on Sròn Dubh with forsteritic and tremolitic limestone and interbedded calcsilicate-rocks.

The sequence between the main and upper limestones on Sròn Dubh consists almost entirely of black schist, rich in pyrite and graphitic in parts, although some 15 to 20 m below the upper limestone there is a unit of pale grey calcsilicate-rock 4 to 5 m thick. Elsewhere this part of the formation is restricted to small enclaves of rusty weathering semipelite and micaceous psammite within amphibolite.

A near-complete section through the upper limestone is present roughly 150 m to the south-south-west of Sròn Dubh north summit. It is 16 to 17 m thick and consists predominantly of a pale grey to white massive calcsilicate-rock within which are two beds, 1 and 3 m thick respectively, of pale to mid-grey forsterite-diopside limestone. Massive white calcsilicate-rock with amphibole porphyroblasts, pale green calc-schist and pale grey banded limestone recorded 150 m north-east of Creag a’ Chlamhain summit within amphibolite may also represent the upper limestone.

The upper limestone is overlain by rusty weathering semipelitic hornfels and coarse-grained amphibolite on Sròn Dubh. The latter rock type seems to occur quite commonly at this level in the succession, a further three pods being recorded within the granite roughly 400 m south-south-west of the upper limestone, and on the south-western slopes of Geallaig Hill [NO 28 97] to [NO 29 97] where it apparently forms the core of a large synclinal fold.

Geallaig Hill is largely devoid of exposure, but the distribution of float suggests that the amphibolite is underlain by a narrow strip of black schist with interbedded calcsilicate-rocks that can be traced almost to the north-east-trending fault crossing the northern end of the Sròn Dubh spur. This black schist is underlain by semipelitic to psammitic schists which are thinly interbedded with laminated quartzose psammites. These in turn pass down into poorly exposed schistose cordierite-rich semipelites and pelites with minor psammites whose outcrop extends northwards to intersect the northern boundary of the district between [NJ 283 007] and [NJ 307 006]. Although black schist is absent, these rocks are regarded as part of the Crathie Formation because the outcrop extends northwards into the Glenbuchat district and is exposed at Delnabo Quarry [NO 304 002], where grey weathering limestone similar to the main limestone is seen.

Lawsie Psammite Formation (QBA)

If the ghost stratigraphy within the amphibolite on Creag a’ Chlamhain [NO 26 95] is correct, the upper limestone of the Crathie Schist and Limestone Formation is succeeded by a dominantly psammitic sequence, the Lawsie Psammite Formation, which is correlated with the Glen Callater Banded Psammite of the Braemar district. The type section of the Lawsie Psammite Formation occurs 200 to 350 m north-east of the summit cairn on Creag a’ Chlamhain where it is separated by 70 to 80 m of amphibolite from the upper limestone and by a 30 m-wide diorite dyke from the overlying Rimarsin Limestone Formation. The formation consists principally of rusty weathering psammite and micaceous psammite which, in places, is quite finely laminated or colour banded. Local lithological variations include the occurrence of massive unbedded psammite with intercalations of dark semipelite at [NO 2687 9582], and psammite and white weathering quartzite within amphibolite on Knock of Lawsie [NO 26 96]. At one locality [NO 2608 9647] the quartzite is characterised by biotitic laminae and yellow-weathering porphyroblasts, probably of staurolite.

Away from the type area, the formation has an extremely restricted outcrop, and is not shown on the 1:50 000 scale map. On the northern slopes of Craig Nordie [NO 23 94] the dioritic rocks enclose screens, up to 250 m long, of rusty weathering psammite and micaceous psammite with subordinate semipelite, which appear to overlie enclaves of black schist; striped micaceous psammites underlie a bed of pale limestone, lithologically and stratigraphically comparable with the Rimarsin Limestone Formation on the eastern slopes [NO 32 98] of Carn Dearg. South of the Dee possible occurrences of the formation are confined to two large xenoliths within granodiorite located approximately 280 m north-north-west of Princess Helena’s Cairn. The larger of these, situated immediately east of a deer fence at [NO 2560 9408], is 3 m across and comprises pale grey micaceous psammite (S77648) with a distinctive striping and prominent micaceous partings which dip steeply to the south.

The main constituent of the Lawsie Psammite Formation is very similar in appearance to the psammite of the Creag nam Ban Formation of the Easdale Subgroup, so in areas where the Rimarsin Limestone and Creag Leacach Quartzite are not present the positioning of the Appin–Argyll group boundary is rather uncertain. Such is the case in the Torgalter Burn area [NO 28 95] to [NO 28 96] where psammites that appear to be a northerly extension of rocks of undoubted Creag nam Ban affinity crop out in close proximity to units of the Crathie Schist and Limestone Formation.

Rimarsin Limestone Formation (bLBA)

On Creag a’ Chlamhain the Lawsie Psammite Formation is overlain by a formation consisting predomiantly of limestone and calcsilicate-rock which equates with the Cairn Aig Mhala Limestone in the Braemar district (Figure 7). The limestone is typically pale grey to brownish white and well bedded. It is distinguished in the field from the Crathie Schist and Limestone Formation main limestone by its tan weathering. The type section [NO 2709 9580] is the only locality where the underlying psammite is also exposed. There, a minimum true thickness of 8 m was recorded, comprising a lower unit of green calcsilicate-rock 1.5 to 2 m thick, 3 m of limestone and an upper green and purple calcsilicate unit of which only 3 m are exposed. The nearest exposure of the overlying Creag Leacach Quartzite Formation is approximately 30 m to the east. Between 50 and 100 m along strike to the north-west the formation thickens to over 12 m and the lower calc-silicate unit includes psammite and coarse tremolitic rock.

At Rimarsin [NO 2631 9650] the exposure of the limestone is small but two shallow depressions to the west (sink holes or trial pits) indicate the more widespread extent of the formation. Blocks of a comparable limestone were found some 400 m north-east of Rimarsin; immediately to the west and presumably underlying the source of the float is a succession comprising interbedded semipelite, fine-grained, pale and dark green, and in places epidotic calcsilicate-rock, and coarse amphibolite.

South of the Dee, the only known outcrop of the Rimarsin Limestone is a pod-like body centred on the northern slopes of Tom Buailteach. The most extensive exposure is in the disused quarry at Easter Balmoral [NO 2758 9381] where a continuous cross-strike section of 12.5 m true thickness through the calcareous rocks is preserved on the south face. This consists of 11.5 m of thinly bedded (20 to 50 mm) pale to mid-grey limestone with disseminated sulphide and some green calcsilicate-rocks and cross-cutting calcite veins. At the eastern end of the face, the limestone is overlain by 1 m of more massive pale bluish green calcsilicate-rock with prominent idocrase porphyroblasts. In the less well exposed northern part of the quarry, limestone representing a lower stratigraphical level than that seen on the south face contains interbedded calcsilicate units, including a very fine-grained white rock of cherty appearance. A unit of fine-grained hornblende schist (0.20 m thick) occurs within the calcareous rocks on the eastern wall of this part of the quarry. The youngest rocks exposed in the quarry lie 70 m north-north-west of the nearest exposure of the Creag Leacach Quartzite Formation and, in contrast to the area of the type section, the intervening rocks are exposed, albeit on a very limited scale. Within 20 m of the quartzite these comprise dark grey cordierite and sillimanite-bearing semipelites which become more pelitic downwards. These rocks overlie the more characteristic calcareous lithologies which here also include an amphibolitic calcsilicate-rock.

Two thin sections of the limestone serve to show the ranges of purity and grain size. The first (S82341), collected from the Rimarsin locality, is somewhat coarser grained (0.75 to 2.5 mm and locally up to 5.0 mm) and purer than the second, comprising 80 to 90 per cent carbonate. The remainder of the rock consists of diopside, quartz, plagioclase, opaque minerals and sphene. In contrast, the second specimen (S81984), which was collected from the south wall of the Easter Balmoral Quarry, has only 50 to 60 per cent carbonate grains which average 0.5 mm across and seldom exceed 1.0 mm. This suggests there is an inverse relationship between purity and carbonate grain size in these rocks. The principal impurities are anhedral to granular diopside, largely sericitised plagioclase, hematite and opaque minerals. Scapolite (up to 2.5 mm across), rare tremolite, idocrase and sphene are also present.

Chapter 6 Dalradian—Argyll Group (1)

Islay Subgroup

Boulder Bed (Z)

Limited outcrop of this important marker horizon was recorded during the primary survey of the western half of the Braemar district (Sheet 65W; Cunningham Craig and Barrow, 1912). Cox (1966) and Upton (1986) extended the known outcrops and added new exposures, and Upton coined the name Sròn na Gaoithe Boulder Bed. In the course of this survey, small exposures and loose blocks of Boulder Bed were discovered at a number of localities on the spur east of Carn Dearg; grid references for the best localities are [NO 3269 9799], [NO 3299 9793] and [NO 3317 9795], although at the last mentioned locality it occurs in the form of loose material considered to be nearly in situ. It is a medium-grained, phyllitic to schistose, micaceous rock containing angular to subangular clasts of granitic rock up to 0.25 m across, together with less abundant amphibolite clasts. The intensity of the deformation in the rock varies considerably between the three localities. Contact relationships with the adjacent lithologies are not exposed, but at the latter two localities the Boulder Bed is encircled by amphibolite of the younger basic suite. At the first locality schistose psammite to semipelite occurs on both sides of the exposed Boulder Bed with extensive development of amphibolite to the north-east. To the south-west these schists give way to quartzite of the Creag Leacach Quartzite Formation within which limited younging evidence (possible cross-bedding [NO 3270 9786] and graded bedding [NO 3338 9748]) indicates that it lies stratigraphically above the Boulder Bed.

Creag Leacach Quartzite Formation (QGI and QSI)

The most extensive exposure of this formation in the Ballater district is south of Crathie where it can be traced nearly continuously for 1.5 km between Tom Buailteach [NO 2761 9375] and the northern slopes of Tom Bad a’ Mhonaidh [NO 285 924]. The lower contact with the Rimarsin Limestone Formation on Tom Buailteach was broadly described in the previous chapter. The boundary with the overlying Creag nam Ban Psammite Formation, as defined by the first appearance of interbedded micaceous psammite and semipelite, can be traced with some certainty for 800 m, from the southern edge of the Khantore Granite [NO 2843 9326] to where it is intersected by the eastern edge of the Lochnagar Granite. The outcrop width in this section is between 400 and 600 m, but the relationship of these figures to true thickness cannot easily be established, because over much of its length the western or lower contact is against granite, and contrarily where the base of the formation is exposed on Tom Buailteach the upper contact is not seen. The wide variation in dip of bedding, from 40º in the east to 78º in the west, and the notable swing in strike from north-north-west in the south, to east–west on Tom Buailteach also hamper an estimation of the true thickness. The outcrop can thence be traced westwards for over a kilometre through a series of diorite-enclosed blocks, within which the bedding dips southwards at between 30º and 40º. The largest of these is 400 m long and occupies the summit of Tom na h-Ola [NO 275 928].

Quartzite blocks appear 1 km north of the Dee, [NO 2728 9553], from where the outcrop of the formation can be traced, by in situ and float material, for 1.7 km to the north-north-west. In this area the outcrop is much more fragmented than south of the Dee, mimicking that of the underlying Appin Group rocks. At no point is the quartzite seen in contact with either the preceding or succeeding Dalradian formations, but the outcrop width is constrained to between 300 and 400 m by isolated exposures of these other rocks. Since the bedding in the quartzite dips quite consistently to the east-north-east at between 65º and 80º, the outcrop width should correspond to a true maximum thickness of 270 to 390 m. No trace of the Creag Leacach Quartzite Formation was found around the nose or on the eastern limb of the Creag nam Ban Syncline for example at [NO 288 962], but the position of the formation can be picked up again in Easter Micras Burn [NO 3012 9665], where small exposures of quartzite and calc-quartzite are present at the contact between the Abergeldie Diorite and the Coilacriech Granite.

Between 2.5 and 3.5 km to the east of this, the roof zone of the Coilacriech Granite is made up largely of amphibolite which contains numerous rafts of quartzite, 20 to 30 m wide (locally up to 100 m) and up to 700 m long. They are mostly parallel and elongated in a north–south direction, swinging to north-east-trending in the south. There is little evidence of bedding in these strips, but the alignment noted above follows the regional strike of the lithological layering. At one locality [NO 326 981] in this area, the quartzite forms the core of a south-east-facing fold, with the passage to Boulder Bed, described previously, on the north-east limb.

The Creag Leacach Quartzite Formation is predominantly made up of a clean, white-weathering, feldspathic orthoquartzite which is pale grey to brown where fresh. It lacks the glassy and consistent appearance of the Cairnwell Quartzite from which it is also distinguished by finer grain size and higher feldspar content. The formation ranges from massive and unbedded, particularly in the north-east part of the outcrop, to well bedded, as defined by alternating quartz- and feldspar-rich layers or, more rarely by heavy mineral bands or micaceous partings. Pebbly beds are also common; individual clasts are dominantly of quartz and up to 100 mm in diameter. Upward-fining graded bedding within these pebbly units was recorded at a number of locations for example [NO 2736 9282] where several units are present. The quartzite consists principally of quartz, plagioclase, K-feldspar and minor biotite.

The formation also contains other lithologies including psammite, semipelite and, more rarely calcareous rocks. These usually take the form of thin beds within the quartzite, but on the southern slopes of Carn Dearg between [NO 327 978] and [NO 336 964] the succession is dominated by a north-easterly tapering wedge of micaceous psammite with little or no interbedded quartzite. Elsewhere the interbedded material is dominantly psammitic; on Creag a’ Chlamhain [NO 2689 9596] this comprises pink weathering striped psammite with finely divided biotite, whereas farther north [NO 2678 9619] and [NO 2701 9611] the rock is characteristically finely laminated.

South of the Dee, instances of other lithologies are much rarer, but in the wooded area 200 m west of Buailteach [NO 2753 9325] the quartzitic sequence contains a notably heterogeneous semipelitic unit. The rock (S81978) is basically a fine-grained, dark grey, cordierite-rich schist which is cut by paler quartzofeldspathic layers and biotite-feldspar laminae. The heterogeneity stems from the flowing of the quartzofeldspathic material between cordierite-rich boudins, which imparts to the rock a pseudobrecciated texture. In this semipelitic unit (for example (S81978)) the cordierite porphyroblasts and poikiloblasts may exceed 10 mm in length, and are notable for their volume of included corundum.

Rocks containing calcsilicate minerals were recorded in association with quartzite at two locations, but in both instances the exposure was small and enclosed by intrusive rock so their position within the formation is not known precisely. The first of these occurs by the path [NO 2636 9372], 100 m south-south-east of Princess Beatrice’s Cairn, and consists (S82183) of a hornfelsed semipelite with pale green calcsilicate layers. This forms part of an assemblage of micaceous quartzite, amphibolite and hornblende schist which is extensively veined by coarse tonalite. The layers (for example (S8213)) comprise diopside and plagioclase with subordinate hornblende and accessory sphene and iron oxide. The other occurrence is on the east bank of the Easter Micras Burn [NO 3012 9665] and comprises a relatively massive, fine-grained, pale greenish grey rock with little obvious bedding. Apart from the colour, this rock (S93079) is practically indistinguishable in the field from the typical quartzite; quartz is the dominant mineral phase, but does not constitute more than about 60 per cent of the rock, and granular diopside makes up 10 to 15 per cent.

Subtle variations in mineralogy and texture within the quartzites are exemplified by two samples from the Tom Buailteach area. One example (S81979) is dominated by irregularly shaped, anhedral plates of strained quartz, up to 1.8 mm across, which lack any evidence of preferred alignment. The grains invariably have sutured boundaries and are locally interlocking. Somewhat smaller equant grains of feldspar occur within or, more commonly on the margins of the quartz. These consist largely of slightly sericitised plagioclase and appear to have been partially resorbed by quartz. Similarly sized grains of K-feldspar are present, but are subordinate to plagioclase. The remainder of the rock comprises tiny flakes of largely chloritised biotite, zircon, opaque minerals and possible allanite.

In the second sample (S81977) the quartz forms somewhat larger (up to 3 mm) grains, some of which are elongated parallel to the foliation as defined by the common alignment of biotite. The inclusion of aligned biotite is an indication that these large quartz grains grew at a late stage in the history of the rock. However, this probably occurred before the thermal overprint since the larger quartz and feldspar grains have locally been totally or partially replaced by finer grained granoblastic polygonal aggregates. As in sample S81979 both plagioclase and K-feldspar are present, predominantly as small grains, although the former also forms clasts up to 1.7 mm. Red-brown biotite occurs principally as discrete flakes, mostly with a common alignment, but may be concentrated into enriched laminae. The rock also includes small plates of secondary muscovite and anhedral grains of partly pinitised cordierite.

Easdale Subgroup

Creag nam Ban Psammite Formation (QSE)

This formation consists largely of psammite and semipelite and can be traced for over 12 km, from within the Lochnagar Granite to the northern edge of the Ballater district. The most southerly outcrops comprise two ovoid screens within Lochnagar L1 Granite, 80 and 180 m in diameter respectively [NO 298 898] and [NO 296 894] on the northern slopes of Meall Gorm. These are the largest known inclusions within the eastern part of the Lochnagar Granite which is, for the most part, characterised by a paucity of extraneous material. The only other sizeable enclaves within the L1 granite [NO 281 905] are also of Creag nam Ban Formation rocks.

Between 0.5 and 2 km north of where the Easter Balmoral–Glen Muick hill track crosses the Girnock Burn [NO 293 909] there are extensive exposures of the formation. On Tom Bad a’ Mhonaidh these take the form of lensoid screens, mostly aligned north-north-west, within granodiorite of the Abergeldie Complex. On the lower ground to the north the metasedimentary rocks predominate, although they are pervasively intruded by veins and sheets of granite and granodiorite. This is the type area for the lower part of the formation and includes the only known exposures of the boundary with the underlying quartzite. In this and the exposures to the south, the lithological layering (which may represent bedding) and the early cleavage are steeply inclined, and strike either north–south or north-north-west.

Around Genechal [NO 290 929] the outcrop of the formation is truncated by the Khantore Granite, and between the northern edge of the intrusion and the River Dee there are only scattered small exposures. North of the Dee, roof pendants and screens of Creag nam Ban lithologies are present within granite on The Maim [NO 277 972] to [NO 279 969] and around Torgalter Burn [NO 287 962] to [NO 288 956]. Between these two areas the outcrop of the formation effectively wraps around the axial trace of the F2 Creag nam Ban Syncline (Figure 16) and veers south-south-east towards Creag nam Ban.

The Creag nam Ban ridge is the type area for the formation (excluding the lowest beds) with a near-continuously exposed vertical section of almost 300 m on the eastern side, and excellent exposures of the upper boundary. On the western flank of the ridge it is clearly evident that the formation forms the roof zone at the eastern edge of the Khantore Granite. The rocks on the ridge also lie within or immediately east of the core of the Creag nam Ban Syncline, so the orientation of the lithological layering shows considerable variation. From the eastern slopes of Creag nam Ban the formation can be traced southwards, and then eastwards round the closure of the Camlet Anticline, before regaining the regional north-east strike in Glen Girnock. To the north-east the outcrop width is progressively reduced to under 100 m by the cross-cutting eastern edge of the Coilacriech Granite. North of the Dee, the outcrop reverts to the more usual pattern of amphibolite- and granite-hosted discrete blocks, the southernmost of which includes the contact with the underlying Creag Leacach Quartzite Formation.

Accurate thickness estimates of the Creag nam Ban Psammite Formation are confounded by lack of continuous cross-strike sections, folding and large intrusions. However, an approximation can be obtained from two sections.

  1. On the north side of the Dee between Coilacriech [NO 328 969] and Balanreich [NO 340 965], the cross-strike width between the top of the Creag Leacach Quartzite Formation and the base of the Glen Girnock Calcareous Formation is about 550 m which corresponds to a true thickness of about 515 m (including 300 m of amphibolite).
  2. The corresponding measurement between the track junction [NO 285 925] and the lower slopes of Sgor an h-Iolaire [NO 299 927] is 1440 m, of which 980 m is across granite, giving an estimated true thickness of 450 to 460 m.

The Creag nam Ban Psammite Formation is dominated by a pale grey to buff micaceous psammite which in many places is characterised by a fine, millimetre-scale lamination. Locally this may be sufficiently pervasive for the rock to be deemed a laminite and visibly resembles a non-graphitic version of the Cairn of Claise Transition (Figure 7). In places (for example (S82047)) the rock may contain slightly coarser, 4 to 8 mm-thick, pale and mid grey colour banding. A typical example of the laminated psammite (S82008) in thin section comprises a granoblastic aggregate of quartz, plagioclase and foxy brown biotite with sporadic poikiloblastic to fragmented garnet. K-feldspar was not positively identified, but some areas comprise quartz and a heavily sericitised feldspar which contrasts markedly with the relatively unaltered plagioclase grains elsewhere in the thin section.

The lamination, which is somewhat discontinuous, results from the presence of thin (single crystal deep) biotite-rich laminae. The additional presence of disseminated iron oxide and zircon would seem to indicate that some, if not all, of these are heavy mineral bands. Cordierite, typified by its concentration of tiny biotite inclusions, also occurs within the more pelitic laminae, but in places is seen to be growing across the psammitic laminae.

Layers of gritty psammite and micaceous psammite, containing pebbles of brown quartz up to 15 mm in size, and slightly smaller white feldspar clasts were recorded [NO 2973 9492] on the northern face of Creag nam Ban. No graded bedding was noted at this locality, but a thin section (S82060) [NO 3017 9413] contains three such units the thickest of which is up to 2.8 mm.

Some compositional and textural differences are apparent in the psammites which make up the lower and uppermost beds of the formation. For instance, the only recorded occurrence of K-feldspar in interlayered semipelite occurs in the lowermost 100 m or so. In the Tom Bad a’ Mhonaidh–Genechal area, the same part of the succession contains beds of white-weathering quartzose psammite which superficially resembles the Creag Leacach Quartzite Formation. However, in thin section (S81990) it is seen to be relatively enriched in biotite and additionally contains sericite-pinite-serpentine pseudomorphs after cordierite, and late muscovite porphyroblasts.

Further examples of relatively mica-poor psammite layers at this position in the succession are to be found north of the Dee [NO 2697 9664] on the spur linking Creag a’ Chlamhain and The Maim. However, the rocks here (for example (S82347)) are slightly more micaceous than those south of the river, although this is not immediately apparent since the biotite is relatively fine grained. Significantly, the biotite in these rocks is pleochroic from yellow to brown and not foxy brown which may reflect a lower degree of thermal metamorphism; the surrounding rocks are granite rather than granodiorite. Because of their restricted stratigraphical occurrence these particularly psammitic units are useful markers in areas of poor exposure or uncertainty. For example, on the south-east side of Sròn Dubh a 5 m-thick unit of laminated quartzose psammite within micaceous psammite [NO 2875 9615] was used, in the absence of outcrops of Creag Leacach Quartzite and Rimarsin Limestone formations, to distinguish the lower part of the formation from the lithologically similar Lawsie Psammite Formation.

Beds of similar mica-poor psammite and quartzite, which are commonly marked by iron staining are locally present near the top of the formation, most notably around Cnap na Cuile [NO 306 921] at the southern end of the Creag nam Ban ridge, and less extensively [NO 298 900] on the eastern edge of the Lochnagar Granite where units up to 7 m in thickness are interbedded with micaceous psammite and semipelite. However, the most extensive development of these rocks is in the area between the Glen Girnock track [NO 318 939] and the Duchess of Kent’s Cairn [NO 299 938], where a lithologically distinctive and strongly deformed, rusty-weathing quartzite and psammite unit, mappable on the 1:50 000 scale, defines the top of the formation. The thickness of the unit is conservatively estimated at 60 m, although in the core of the Creag nam Ban Syncline the outcrop width extends to nearer 200 m. The most complete section seen through the unit is at [NO 313 937] where the base is marked by a 10 m-thick unit of clean white quartzite with interbedded well-cleaved mica schist. Above this, it consists predominantly of laminated micaceous psammite interbedded on a centimetre to metre scale with quartzite to quartzose psammite and minor semipelite (Plate 4).

Close to the axial trace of the syncline [NO 300 938], the unit includes a thin layer of hornblende schist with thinly interbedded quartzite which more closely resembles rocks at the base of the overlying Glen Girnock Calcareous Formation.

Locally, the psammites contain semipelite units interbedded on a centimetre to metre scale. They are compositionally and texturally more varied and considerably more heterogeneous than the psammites. They show the effects of the widespread thermal metamorphism, particularly through the partial destruction of the regional planar fabric and the dark grey coloration imparted by the ubiquitous cordierite. A thin section (S82012) that typifies many of the rocks examined, consists essentially of quartz, plagioclase, biotite, cordierite and minor andalusite with accessory iron oxide, zircon and apatite. In this rock, quartz and, to a lesser extent, plagioclase tend to occur as discrete lenses or discontinuous laminae. Plagioclase also forms granoblastic aggregates with biotite and altered cordierite. Red-brown biotite is ubiquitous throughout the rock, mostly as stumpy crystals, but more rarely as larger plates, which are invariably aligned parallel to a layering largely defined by concentrations of biotite with or without altered cordierite. This specimen also serves to illustrate the great heterogeneity of these rocks possessing the following domains:

Fibrolite, hercynite and corundum may also be present in these rocks.

The Creag nam Ban Formation is largely devoid of amphibolitic rocks, but at three localities [NO 3024 9480], [NO 3021 9467] and [NO 39 9448] on the eastern side of the Creag nam Ban ridge thin units of cummingtonite-bearing schist, ranging in thickness from 15 to 40 mm, are contained within the psammite–semipelite succession. The rock is dominated by 3 to 4 mm-long sieved porphyroblasts of pale green to colourless cummingtonite and smaller (1 to 2 mm) garnet porphyroblasts. The cummingtonite-bearing schist is characterised by abundant apatite, which is reflected in the unusually high P2O5 of the two analysed rocks (Table 3).

Glen Girnock Calcareous Formation (bCE)

The position of the Glen Girnock Calcareous Formation within the Easdale Subgroup is analogous to that of the Ben Lawers Schist of the Pitlochry district (Figure 7). In broad terms there are also many lithological similarities between the two formations, notably the high calcareous content, the presence of amphibolite sheets including probable metabasites, and of pyritic schist. However, whereas the Ben Lawers Schist is dominated by calcareous pelite the Glen Girnock Formation includes a significant proportion of more clastic and non-calcareous rocks which reflect a greater input of terrigenous material into the basin.

Within the Ballater district the formation has a strike length of just over 13 km. It can be traced almost continuously between the Lochnagar and Ballater granites, and thereafter intermittently through amphibolite-hosted screens to the northern edge of the district. Throughout much its length the formation strikes north-east. More variation is apparent in the Ballater area where certain sections show a more east-north-east-orientated alignment, and close to the northern edge of the district the general trend is north-west. South-west of Ballater, the only significant variation occurs in the area around Camlet [NO 308 931] where the alignment of the basal units is affected by a major F2 fold pair.

The most completely exposed section through the formation (hence the type section) is between a point 150 m south-east of Littlemill [NO 326 955] and the bend in the forestry fence about 500 m north of Creag Liath [NO 337 942]. Within this section the outcrop width is about 1.7 km which corresponds to a true thickness of 1.2 to 1.3 km. Traced to the south-west, the outcrop width is gradually reduced and in the ground south-east of Bovaglie, close to where the formation is truncated by the Lochnagar Granite, it measures only 950 m which translates to a maximum true thickness of 820 to 890 m. This change in thickness stems largely from a south-westerly reduction in the volume of interbanded amphibolite. The formation as a whole is characterised by a large variation in unit thickness, most evident in the outcrop pattern of the Knock Semipelite Member and Creag Liath Pelite members, but this thickness variation can also be seen in individual beds. It is probable that this is in part tectonic — the upper part of the formation is within the Coyles of Muick Shear Zone — but it is evident elsewhere that this is the result of lateral facies variations (Figure 10).

The Glen Girnock Calcareous Formation is dominated (Figure 10) by an assemblage of interlayered amphibolitic, calcsilicate- and semipelitic rocks. However the succession also includes limestone, psammite, quartzite and highly micaceous pelite and, as a whole shows significantly greater lithological variation than any other unit in the Dalradian of the Ballater district. The formation can be divided into six lithologically distinct members, although only the Knock Semipelite and Creag Liath Pelite members are sufficiently distinctive in the field to be delimited on the 1:50 000 Series map. The members are described below in ascending lithostratigraphical order.

Basal Mixed Member

The lowermost member of the formation is characterised by a wide variety of lithologies which include semipelite, hornblende schist, quartzite, psammite, calcsilicate-rock and limestone. In the type section [NO 3264 9566] this member is less than 150 m thick. The base is well seen at the type section where relatively homogeneous semipelite marking the top of the Creag nam Ban Psammite Formation is overlain by more diverse lithologies dominated by laminated micaceous psammite, but including interbanded hornblende schist, semipelite and quartzite. A unique feature of the psammite (S94395) is the presence of a few ragged anhedral cordierite porphyroblasts, up to 3.3 mm long, which are aligned along an incipient crenulation cleavage at an angle of 50° to the dominant schistosity. A highly schistose andalusite-cordierite semipelite collected from near the base of the member (S94936) is also unusual in that biotite is extremely rare, being confined to within andalusite, the pelitic microlithons containing muscovite and subordinate chlorite.

Between 40 and 50 m above the base the succession becomes distinctly more calcareous, being dominated by a flaggy, fine-grained, pale to dark grey calcsilicate-rock with interbedded tremolite schist, actinolite schist and rare quartzite. Above 110 m from the base more massive hornblende schist or amphibolite units make their appearance within the calcareous rocks marking the passage to the overlying dominantly metabasic assemblage. Subtle differences, possibly reflecting differing protoliths, were noted in the two actinolite schists examined. (S94400) is a dark greenish grey schistose rock which resembles a fine-grained version of the meta-igneous type hornblende schist. In fact, it contains over 70 per cent pale green actinolite, mostly in the form of a granoblastic felt but within which are sporadic slightly larger prismatic grains. Fresh granoblastic plagioclase and finely divided iron oxide are scattered through the rock. In one place the felt is broken by an elongate lens of a much coarser grained assemblage of heavily sericitised plagioclase, actinolite and clinozoisite.

The other example (S94398) is a somewhat coarser intergrowth of distinctly prismatic actinolite which constitutes probably up to 80 per cent of the rock. Interstitial granoblastic plagioclase is quite strongly sericitised and is intergrown with quartz which also occurs poikilitically within actinolite. Sphene is also present as scattered small grains. The most significant difference however, lies in the presence of pyrrhotite which is scattered throughout much of the rock and locally concentrated in selected areas such that the rock has a magnetic susceptibility of 1.4 SI. The fact that pyrrhotite is also common in adjacent tremolite schist may be an indication that this actinolitic rock has a sedimentary protolith.

Pyrrhotite may also be present in rocks at the base since the member can be distinguished magnetically from both underlying and overlying rocks. The magnetic signature can be utilised to trace the subcrop of the member through the largely drift covered ground around the River Dee, as far as the Bridge of Gairn Fault. Beyond the fault the member can be traced north-eastwards for a further 3 km to the western edge of the Ballater Granite. On Craig of Prony [NO 352 988] and on Hill of Candacraig [NJ 346 001] rocks of the member are present as slivers within amphibolite and comprise semipelite, calcsilicate-rocks, including tremolite schist, and hornblendic rocks with scattered limestones, some of which are forsteritic. Pyrrhotite/pyrite is widely disseminated, but may also be sufficiently concentrated as to form a mappable unit (see Chapter 2) which can be readily identified by its rusty weathering.

In the area between Bovaglie [NO 30 92] and Loinveg [NO 31 93] the base of the formation is locally marked by about 15 m of fine-grained and generally well-cleaved hornblende schist. The schist is interbedded with clean white quartzite mostly on a metre scale, although much thinner layers occur to a limited extent in the lowermost unit. This gives rise to a visually very distinctive lithology that forms a marker horizon, but for only a very limited lateral extent. Elsewhere in the Dalradian such fine-scale interbanding with metabasic rocks has, been considered to indicate a possible volcanic origin. If this is so here, it is possible that the quartzite, which is invariably fine grained and of high purity, originated as a sinter type chemical sediment; this might go some way to explaining its lack of lateral persistence. Elsewhere in the district the base of the formation is marked by semipelite which is distinguished from that in the underlying Creag nam Ban Psammite Formation by abundant quartz segregations.

Lower Metabasic Member

The sequence above and below the Upper Mixed Member is dominated by hornblendic metabasic sheets with subordinate interlayered metasedimentary rocks. There seems little doubt that the bulk of the sheets are metadolerites which were intruded into the sedimentary rocks prior to the second and possibly even the first deformation. As such they will be described in more detail in Chapter 9. There is also the possibility that, like parts of the Ben Lawers Schist in Perthshire the formation has a volcanogenic component. Probably the most convincing evidence for this is to be found on the south-facing slopes of Hill of Candacraig where fine-grained schistose metabasic bodies are interlayered on a centimetre-scale with plagioclase-tremolite schists. Taken together with apparent transitions between the two lithologies and a similar recognisable structural history, there seems little doubt that these metabasic rocks formed at about the same time as the metasedimentary rocks and as such have volcanic or pyroclastic protoliths. Also, in the Polhollick area [NO 344 965] the hornblende schists are closely cleaved and locally epidotic, features which are also common in the Meall Dubh Metabasite Formation. Likewise, the metasedimentary rocks have several unusual features (for example their fine grain size and some relict larger grains) which might be explained by considering them as tuffaceous rocks.

The interlayered metasedimentary rocks are characteristically fine grained; they comprise actinolitic schists, calcsilicate-rocks and schistose semipelites. The actinolitic schists are mainly mid to dark purplish grey in colour and in places may be hard to distinguish from the finer grained hornblende schists. They consist principally of pale green to colourless actinolite and plagioclase with subordinate chlorite, sphene and iron oxide. Rare porphyroblasts of idocrase were recorded in (S82051). The variation in form of the amphibole suggests that it has replaced hornblende and/or diopside, although no precursors have survived.

The calcsilicate-rock components exhibit a wide range of colour and composition. Typical of the most widespread type (S82195) is a pale greenish grey banded rock composed of plagioclase, quartz, epidote, pale green fibrous amphibole and sphene. This rock occurs as thin layers within semipelite with amphibole generally concentrated along the boundary between the two lithologies. Less common amongst the calcsilicate-rocks are those in which a dark grey coloration reflects a high concentration of finely divided opaque minerals ((S99440) and (S82845)). Conspicuous iron staining on weathered surfaces (S99440) coupled with low magnetic susceptibility, suggests that the opaque phase includes pyrite. A well developed slaty cleavage is another distinguishing feature of this rock which comprises an extremely fine-grained granoblastic intergrowth of quartz, plagioclase, very pale brown phlogopite, tremolite, opaque minerals and sphene. Diopside occurs locally with much coarser grained quartz.

Knock Semipelite Member (SE)

The Knock Semipelite Member has a strike length of only 3 km. It reaches its greatest development on The Knock [NO 34 95], a hill about 2 km due west of Ballater, where the outcrop exceeds 300 m in width. However, it is impossible to relate this figure to actual thickness, principally because the upper contact is nowhere seen and the dip of bedding cannot be ascertained in the lithologically uniform sequence. The member extends for a short distance north-eastwards to the Bridge of Gairn Fault, beyond which it is cut out by the Ballater Granite. Immediately south-west of The Knock the member is largely concealed, but small exposures of the adjacent lithologies allow a maximum outcrop width of less than 200 m. The first conclusive evidence that the member thins to the south-west occurs within the type section [NO 3304 9606], where the outcrop width is less than 40 m. The outcrop is poorly exposed, but coincides with a small gully which can be traced to the south-west for about 300 m towards a further exposure of the member, some 50 m south of Creag Phiobaidh summit, where a minimum width of 15 m was recorded. Between 200 and 300 m south-west of this, the Lower Metabasic Member is overlain by mica schist which contains less muscovite than to the north, but is nonetheless regarded as part of the Knock Semipelite Member. No trace of the member was found beyond this.

The member is composed almost entirely of a very distinctive muscovite-rich schist which makes it readily identifiable in the field. The only comparable lithology in the area is the Creag Liath Pelite Member which occurs at the top of the formation. The boundary with the underlying hornblende schist, which is exposed almost continuously for over 200 m on The Knock, is always sharp and there is no evidence of interbanding between these two highly distinctive rock types.

The muscovite schist rock is dominantly semipelitic but includes some pelitic and rare psammitic units. It comprises a highly schistose medium- to coarse-grained rock, mid-grey on sections normal to the schistosity, but distinctively silvery grey on cleavage surfaces, reflecting the high muscovite content. The form of the muscovite in places (for example (S82193)) suggests possible derivation from sillimanite. The mineralogy is quartz, plagioclase, muscovite and chlorite that has largely replaced the earlier biotite.

The overall homogeneity of the rock is locally broken by the presence of pale grey quartz or quartz-feldspar lenticles. The heterogeneity is enhanced in thin section (S82193) by the additional presence of pelitic layers or partings which are mostly found adjacent to these quartz-feldspar lenticles. The occurrence of tourmaline, zircon and iron oxide in the pelitic units suggests they are relict heavy mineral bands. The core of the pelite units is a felt of muscovite, chlorite, sericite and/or biotite, scattered through which are isolated crystals of garnet, staurolite, andalusite and rare fibrolite.

Upper Mixed Member

The Knock Semipelite and the Lower Metabasic Member (Figure 10) are overlain by a poorly exposed sequence of interbedded psammite, semipelite, pelite and calcsilicate-rocks which can be traced intermittently south-westwards from the Bridge of Gairn Fault for 5 to 6 km. In the type section, on Creag Phiobaidh, the outcrop width is nearly 500 m.

The northernmost exposures on Craig of the Knock contain little evidence of calcareous rocks; they consist of semipelite with interbedded psammite and rare pelite. The semipelite here is locally coarse grained to gneissose with abundant quartz lenticles. Since these exposures lie within the thermal aureole of the Ballater Granite, the rock is characterised by the presence of cordierite, andalusite and corundum, with garnet and staurolite only rarely seen. Fibrolite and K-feldspar were additionally noted in (S82190).

In and around the type section on Creag Phiobaidh, the member is made up of a more varied assemblage of lithologies, particularly in the lower half, including semipelite, pelite, psammite and calcsilicate-rock. The sequence is dominated by fine-grained, dark grey semipelite which gives rise to distinctive slabby faces. The rock consists typically (S94419) of a fine granoblastic aggregate of quartz, dark brown biotite, garnet, chlorite and traces of calcite. Variations within this rock include bands or lenses of slightly coarser grained and more pelitic rock, and coarser grained domains of quartz, ragged garnet, hornblende and sericitised plagioclase. Elsewhere, a slightly more psammitic rock (S94420) includes millimetre- to centimetre-scale bands of coticule (quartz-garnet rock) within which garnet may form up to half of the rock. Most of the interbedded psammite is fine grained and dark grey, but within the lower part of the succession are units of more typical medium- to coarse-grained, pale grey to buff micaceous psammite. This part of the member also locally includes quartzite, mostly as lenses or pods or, more rarely as continuous beds. The calcsilicate-rocks form centimetre- to metre-scale layers of pale grey and green rocks, in places garnetiferous and with ribbed or carious weathering.

South-west of Creag Phiobaidh, these rocks pass laterally into an assemblage dominated by calcsilicate-rocks that can be traced along strike for about 1 km. The outcrop is distinctly lensoid, steadily widening from less than 20 m in a gully 200 m east-north-east of Creag Phiobaidh summit to over 200 m on Tom an Lagain [NO 325 943]. On the west side of the Tom an Lagain Fault, as the top of the Lower Metabasic Member (which can be traced magnetically through poorly exposed ground) swings from south-west to south-south-west so the outcrop of the overlying calcareous rocks narrows rapidly and appears to die out north of the Glen Girnock Fault. At the north-east end of the outcrop the calcsilicate-rocks are separated from the underlying metabasic sheet by the south-western extremity of the Knock Semipelite Member. The upper contact of the calc-silicate assemblage is less readily defined since there is some lithological overlap (interfingering) with the overlying rocks. The most obvious line, that which separates dominantly calcareous from dominantly detrital metasedimentary rocks, is quite oblique to the strike of the lithological layering both on a regional and local scale. Taking into account the broad geometry and limited extent of the outcrop the most likely explanation is that the calcareous rocks originated as a mound or reef (Figure 10). The calcsilicate assemblage consists of a fine-grained, pale greenish grey rock with discontinuous, thin, dark grey micaceous laminae and pink grossular porphyroblasts. A fabric, ranging from a locally penetrative, closely spaced, layer-parallel cleavage to a poorly defined mineral alignment is present in most rocks; more rarely the rock is more massive and flinty looking. The rocks commonly exhibit ribbed or carious weathering. Diopside is the dominant constituent of these rocks, forming up to 70 per cent. It shows a great range in size and in form, from granoblastic to porphyroblastic, and is intergrown with or included in grossularite. Fine-grained granular plagioclase and quartz tend to form widely distributed discrete inclusions, whereas phlogopite, which seems to have replaced diopside, is locally concentrated in lensoid mats.

These calcsilicate-rocks are readily distinguished from those in the Lower Metabasic Member, principally by the widespread occurrence of diopside, the absence of actinolitic schist, and the presence of a thin limestone unit near the base. Exposure of the limestone is confined to two disused pits, although limestone blocks were recorded in the gully at the northern extremity of the outcrop. The larger of the workings is situated 150 m south of Creag Phiobaidh summit [NO 3370 9472], and is 15 m long, 4 m wide and 2 m deep; a well defined, purpose-built cart track indicates a serious attempt to quarry the limestone (Plate 3). As seen in the northern wall of the pit the unit is up to 3 m thick and dips at 45º to the south-east. It (S94421) comprises a pale grey crystalline calcite-rich rock which includes domains enriched in calcsilicates, most notably tremolite with subordinate diopside, plagioclase and, locally phlogopite that is partly replaced by clinochlore. Sphene and pyrite are accessories. The limestone is overlain by massive-weathering pale grey calcsilicate-rock (S94412) which untypically carries porphyroblasts of clinozoisite (enclosing granular diopside), and is cut by hair veins of plagioclase. The footwall comprises a 2 to 3 m-wide felsite dyke.

The second pit, located on Tom an Lagain 250 m south-west of the summit [NO 3236 9412], is no more than a trial scrape possibly reflecting the greatly reduced thickness (1 m) of the unit. The fine-grained calcsilicate-rock which dominates the pit encloses a lens of a much coarser grained, spotted, pale pinkish grey and green diopside-grossular rock characterised by very distinctive carious weathering.

The calcsilicate-rocks above the limestone, when traced to the north-east, show a marked increase in the amount of interbedded non-calcareous material marking a transition to the dominantly detrital part of the member. These interbeds include metre-scale units of laminated micaceous psammite, or, more rarely quartzite, but mostly comprise centimetre- to metre-scale interlayering of fine-grained, dark grey, schistose semipelite to pelite. At one locality very close to the top of the unit, there is evidence of a more intimate and irregular intermingling of the calcareous and clastic lithologies and the possible presence of intraformational breccias ((Plate 5)a). These fragmental rocks are not always apparent in the field, partly as a result of the overall fine grain size of the rocks, but mainly because the compositonal and grain-size differences between clasts and matrix is not always significant, except where calc-silicate clasts are present.

The fragmental texture of these rocks is much more apparent on the cut face of (S94410). It consists of rounded to subangular clasts up to 30 mm in diameter some of which are recognisable as calcsilicate-rocks and others as quartzose psammites. The majority, however are extremely fine-grained siliceous-looking rocks which range in colour from pale to dark greenish grey and may have a foliation that is enhanced by goethite. In places the clasts are supported by dark grey to black matrix which is generally endowed with an abundance of pale pink pin-head sized garnets. An interesting aspect of the garnets is that in places they form a rim to the clasts ((Plate 5)b). The orientation of inclusion trails within these garnets tends to be variable, but significantly seems to follow the outline of the clast. The matrix of this rock comprises a fine-grained aggregate of biotite, granoblastic quartz and garnet.

A unit of magnetite-rich psammite which occurs some distance above the calcareous assemblage can be followed magnetically for up to 4 km along strike, largely through areas of poorly exposed ground, and effectively establishes a link between the succession of the type section with that in the Camlet area. The unit also emphasises the impersistent character of the underlying rocks since, in the type section it occurs close to the top of the member whereas, to the east of Camlet it lies close to the top of the Basal Mixed Member. The magnetically defined outcrop width varies from 50 to 90 m.

Outwardly the rock (for example (S82074)) is little different from other micaceous psammites within the member; it is mid-greenish grey in colour and finely laminated in part. However, it has a magnetic susceptibility of up to 44 x 10−3 SI which compares with less than 1 for psammites elsewhere in the district. The rock consists essentially of quartz, muscovite, biotite, magnetite, plagioclase, sericitised cordierite and accessory tourmaline. Magnetite takes the form of euhedral grains which are distributed throughout the rock with little obvious concentration. They are generally less than 0.1 mm in size, although slightly larger grains are present in micaceous partings.

Upper Metabasic Member

As a whole, the metabasic rocks of this member are similar to those forming the Lower Metabasite Member, but the interbedded metasedimentary rocks contain a much greater proportion of micaceous psammite and semipelite. The proximity of the member to the Coyles of Muick Shear Zone is reflected in the localised appearance of a streaky texture in amphibolites and flattened quartz lenticles in the metasedimentary rocks.

Creag Liath Pelite Member (PE)

From the northern slopes of Creag Liath [NO 34 94] to the valley between Meall Dubh and Craig Megen [NO 31 91], the Glen Girnock Calcareous Formation is topped by a pelite unit noted for its high muscovite content. Like the Knock Semipelite Member, it is sufficiently distinct to be treated as a separate entity on the 1:50 000 map. Evidence from the ground magnetic survey suggests that at its south-western end the outcrop is truncated by a east-north-east-trending fault; certainly it is absent from the well-exposed section at the northern tip of the Craig Megen ridge where the metabasic and metasedimentary assemblage is directly overlain by hornblende schist of the Meall Dubh Metabasite Formation. There is also magnetic evidence that, traced northwards from Creag Liath, the member is progressively cut out by the Coyles of Muick Shear Zone. The outcrop width ranges from 200 m on Creag Liath to 350 m on Meall Dubh, approximating to thicknesses of 150 to 180 m.

After the Knock Semipelite Member this is the most lithologically consistent part of the formation, being dominated by a notably fissile, silvery grey, muscovite-rich pelite. The rock (S94428) comprises a felted intergrowth of aligned muscovite, with subordinate red-brown biotite, cordierite, andalusite, quartz and minor plagioclase. Locally muscovite is subordinate to biotite and the rock may be more semipelitic than pelitic. One thin section (S94433) additionally contains traces of kyanite, the only recorded occurrence of this mineral on the north-west side of the Coyles of Muick Shear Zone. Quartzofeldspathic lenticles are invariably present and show evidence of flattening and boudinage.

Throughout much of the outcrop the only significant lithological variation is the sporadic presence of interbedded micaceous psammite. However, on the spur due west of Meall Dubh summit, semipelite and pelite in the lower part of the member contain an appreciable volume of interbedded quartzite and psammite. The upper part of the spur is dominated by typical Creag Liath pelite, but includes one unit, up to 40 m thick, of hornblendic rocks. The lower section of this unit consists of a finely laminated hornblende-plagioclase schist (S82278), within which are millimetre-scale pale green layers composed of sericitised plagioclase, diopside and sporadic coarse-grained quartz. The rock (S92923) forming the upper section may display a distinctive garbenschiefer texture with randomly orientated poikiloblasts of hornblende in a granoblastic aggregate of quartz, yellow/brown biotite and clinozoisite.

Meall Dubh Metabasite Formation (BE)

The amphibolitic rocks overlying the Glen Girnock Calcareous Formation have a number of features, including finely interlaminated quartzose and epidotic layers, which distinguish them from the more widespread metadolerites (Chapter 10) and which, furthermore indicate an extrusive origin. Volcaniclastic rocks are present locally; these are lithologically similar to the Farragon Beds of the Pitlochry district and occur at a similar stratigraphical horizon.

The outcrop of the formation can be traced magnetically for 6 km, from the eastern edge of the Lochnagar Granite to the point [NO 343 944] where it is cut out by the Coyles of Muick Shear Zone. However, exposure is limited to 4 km of strike between the south-east side of Creag Liath and the northern slopes of Craig Megen. The outcrop is at its widest in the south-west, reaching up to 700 m on Craig Megen. From there it thins to the north-east, being only 400 m on the southern summit of Meall Dubh, and less than 50 m on Cairn an Bealaidh [NO 33 93]. From the attitude of the contacts with the enveloping metasedimentary rocks and some evidence of internal lithological layering, it is inferred that these rocks are steeply inclined, hence the outcrop thickness is only marginally less than the width of the outcrop. The tapering of the outcrop may reflect the effects of the Coyles of Muick Shear Zone, but the presence of interbedded psammite and pelite on the south side of Carn an Bealaidh could indicate a north-easterly thinning of the volcanic pile.

The formation consists largely of well-cleaved, medium-grained dark green hornblendic rocks which invariably contain fine quartz and quartzofeldspathic lenticles, and more rarely exhibit pale and dark green colour banding or streaking. Randomly orientated hornblende crystals are visible on cleavage surfaces. The formation is magnetically anomalous, susceptibilities being in the range 2.5 x 10−3 SI.

The predominant rock comprises a felt of granoblastic sea-green hornblende, which encloses granoblastic to porphyroblastic plagioclase and disseminated magnetite. The freshness and colour of the hornblende is markedly constant throughout the formation; only minor differences are noted (for example (S82289)) where relict brown hornblende cores have survived, and (S82281) where partial replacement by biotite has occurred. The plagioclase shows varying degrees of sericitisation, the coarser grains are invariably more altered and locally the plagioclase has been partly replaced by epidote. It is the relative proportion of plagioclase and hornblende which gives rise to the colour variations noted above.

Movement along the Coyles of Muick Shear Zone is responsible for the omnipresent, closely spaced, penetrative cleavage and the stretching, flattening and boudinaging of more competent units. However, presumably as a result of superimposed thermal metamorphism, this high strain is seldom reflected in the microtextures.

Pale yellow-green epidotic bands and pods, ranging in thickness from a few millimetres to several centimetres are a characteristic feature of much of the succession. A typical example (S94435) comprises a granoblastic aggregate (in part interlocking) of pale yellow pistacite, scattered through which are commonly aligned pale green hornblendes and minor quartz. The total absence of plagioclase in this rock would seem to indicate that it formed the precursor of the epidote. The rock is cut by dark green veinlets, some resembling tension gashes, filled predominantly with quartz, but containing minor hornblende and epidote.

There is little doubt that the rocks described above have an entirely igneous protolith. However, in places (for example (S94450)) the rocks additionally include quartz which is thought to be of detrital origin. The quartz is largely confined to the more feldspathic layers where it occurs either as isolated grains within the plagioclase-hornblende aggregate, or as lensoid aggregates of slightly coarser grains. These quartz-enriched layers are normally of millimetre-scale, but on the northern slopes [NO 313 906] of Craig Megen a discrete unit up to 20 m thick is present. It comprises a pale grey psammite with micaceous partings, and is notable for its anastomosing and attenuated hornblendic lithons. The rock (S82288) comprises three distinct compositional and textural domains.

  1. A relatively fine-grained granoblastic aggregate of quartz, plagioclase, biotite and chlorite.
  2. Essentially as 1 but with the addition of hornblende. The hornblende is present as single small grains and as coarse aggregates whose distribution is quite erratic. Domains 1 and 2 are separated by a thin layer which is enriched in iron oxide and pinitised cordierite.
  3. Coarser grained and more heterogeneous aggregate of hornblende, plagioclase and quartz. Hornblende is largely aggregated and forms discontinuous layers or spots which in places enclose pale pink-brown pleochroic epidote.

A further specimen (S82887) collected from the unit, essentially a plagioclase-quartz-chlorite-hornblende schist, displays two features of particular interest. The first is the presence (albeit very locally) of euhedral epidote grains up to 1.2 mm in length. The mineral has a similar colour and pleochroism to that described above, and additionally shows incipient intergrowth with, or replacement by, pale yellow pistacite. It is thought to be manganian zoisite (thulite). The second point of interest centres on the origin of the hornblende porphyroblasts occurring in lensoid chlorite-free quartzofeldspathic domains within the rock. These may be up to 2 mm long, mostly occur as single grains and are poikiloblastic to almost skeletal. The form is very reminiscent of the cummingtonite in the Creag nam Ban Psammite Formation. The presence of these isolated clots of basic material, in what appears to have been a reasonably mature psammitic rock is somewhat anomalous and is most easily explained by regarding them as the product of a basic ash fall.

Unequivocal volcaniclastic rocks are present on Craig Megen, approximately 90 m south-east of the psammite unit. The rock is particularly striking, partly because of the colour contrast between clasts and matrix and partly because of the extreme stretching of the clasts (Plate 6). Length to breadth ratios are in the order of 2:1 and 3:1 and some clasts may be in excess of 10 cm in length. Three clast lithologies were recorded in thin section (S82295). Dominant are dark green granoblastic hornblende-plagioclase aggregates enclosing heavily sericitised plagioclase and diopside. Less common are coarse intergrowths of poikiloblastic diopside and sericitised plagioclase, carrying subordinate quartz and hornblende. The third category is composed of relatively coarse-grained quartz and hornblende with minor plagioclase and is almost entirely rimmed by hornblende. The matrix resembles the psammite described above, being a pale grey aggregate of quartz, plagioclase, hornblende and minor diopside. A thin section (S82300) collected from the same general locality, shows a similar streaky texture although the clasts are seldom more than 4 cm long. The rock has a distinctly more hornblendic matrix which in contrast to thin section (S82295) is more melanocratic than many of the enclosed clasts. Clast types encountered in this rock include:

Chapter 7 Dalradian—Argyll Group (2)

Crinan Subgroup

The Crinan Subgroup in the Ballater district consists of one formation, the Queen’s Hill Gneiss Formation, which Read (1928) recognised as the lithostratigraphical equivalent of the Ben Lui Schist of the Pitlochry district. Read’s nomenclature is adopted here and extended to cover equivalent rocks to the south-west of Glen Doll, previously referred to as Duchray Hill Gneiss by Barrow and Cunningham Craig (1912) and on 1:50 000 Series Sheet 65W Braemar. The name Duchray Hill Gneiss is retained for the middle member of the Ben Lui Schist Formation in the Glen Shee district (Crane et al., in press).

Queen’s Hill Gneiss Formation (SP and Sc)

The Queen’s Hill Gneiss Formation can be traced along strike for nearly 20 km from the southern edge of the Ballater Granite to the Glen Doll Fault. To the west of the fault it reappears on the southern edge of the Lochnagar Granite and underlies much of the Caenlochan plateau (the upland area in the extreme south-west of the district). Exposure ranges from moderate to good in the north-east, to poor in the central section where much of the upland slopes are blanketed by thick peat. For similar reasons it is sparsely exposed on the Caenlochan plateau, although there are good sections on the fringing cliffs above Glen Doll.

The greatest outcrop width occurs in the north where it exceeds 6 km, but is progressively cut out by the Lochnagar Granite so that where it is intersected by the Glen Doll Fault between [NO 305 825] and [NO 290 788] it is less than 4 km wide. In the northern part the width of outcrop is greatly accentuated by F2 folding. The structural base of the formation, which lies within the Coyles of Muick Shear Zone can be traced for over 3 km between Carn a’ Bhealaidh [NO 330 933] and Craig Megen [NO 316 897], and is well exposed over the summit of Meall Dubh [NO 325 920]. The basal contact is sharp with no evidence of interbanding between the gneisses and the underlying metabasites. It is invariably near-vertical and parallel to the foliation in the gneisses. To the south-west of Craig Megen the contact with the Meall Dubh Metabasite Formation can be traced magnetically to the eastern edge of the Lochnagar Granite, whereas to the north-east of Carn a’ Bhealaidh the magnetic contrast between the two formations is blurred by proximity the Coyles of Muick basic–ultramafic suite.

The Queen’s Hill Gneiss Formation comprises a mixed unit of metasedimentary and metabasic gneisses, and includes a succession of probable volcanogenic rocks which are described separately as the Balnacraig Metabasite Member.

Metasedimentary components

The metasedimentary rocks are predominantly gneissose semipelites and pelites with subordinate psammite and calcsilicate-rocks. Overall there is considerably more lithological variation in the northern part of the outcrop, with the additional presence of quartzite, psammite, calcsilicate-rock and limestone, although the latter two are never very extensive or pure. There is a subtle south-westerly lithological change in the Round Hill of Mark–Allt Darrarie area [NO 33 82], from the mixed sequence to a more monotonously pelitic or semipelitic assemblage; some psammites are present around Capel Burn to the south-west. The change takes place across a north-west-trending lineament which is thought (Chapter 4) to represent a sub-basin margin. West of the Glen Doll Fault the formation consists generally of gneissose semipelite with subsidiary gneissose pelite, and micaceous and quartzose psammite. Psammites, together with lenticular calcsilicate-rock layers, are more abundant locally in the upper part of the formation, marking the transition to the overlying Dounalt Limestone Formation.

The formation is typically migmatitic, particularly where semipelite is dominant. The nature, distribution and origin of the migmatites will be discussed more fully in Chapter 11. In summary, migmatisation produces a crude segregation into discontinuous biotite-rich and quartz-feldspar-rich layers resulting in a diffuse gneissose banding. This largely reflects an earlier compositional layering which is commonly preserved in the hinges of F2 folds, for example near Creag Dearg [NO 35 88]. The compositional layering is a composite fabric because, in places, it contains intrafolial isoclinal folds. It is therefore a transposed rather than a true bedding. The fabric in most rocks is subparallel to both compositional layering and migmatitic segregations.

The gneissose semipelites are generally coarse-grained, strongly foliated (Plate 7) and locally quite heterogeneous. In the northern part of the outcrop, the gneissose semipelite displays a variety of mineral assemblages. The simplest of these consist of quartz, andesine, K-feldspar and dark brown biotite, with accessory opaque minerals, apatite, zircon and sphene. The foliation in the rocks is defined by the orientation of biotite plates, and by elongation of quartz grains, which show considerable internal strain. The feldspar may also have an elongate shape, with extensive marginal polygonisation. Almandine garnets, typically full of biotite and quartz inclusions, are common in pelitic rocks. The average composition is Alm73Py16Gr7Sp3, but zoning is evident with the rims being slightly more Fe-rich and Mg-poor than the core. In places, the garnets have an ‘atoll’ shape where the central part is replaced by a fine biotite symplectite.

Sillimanite is present in gneissose pelitite as fibrolite, and is also found at the centre of atoll garnets; the sillimanite sworls seen in some samples may be pseudomorphs after garnet. Within the formation there are some highly aluminous pelites with up to 30 per cent sillimanite and large atoll garnets set in a biotite-quartz-feldspar aggregate. The lithology is particularly evident around Creag Dearg spur [NO 350 886], where abundant sillimanite and large zoned garnets give weathered surfaces of the rock a very distinctive appearance. In this rock, and more rarely elsewhere, for example Creag Mullach [NO 412 952], the sillimanite forms bladed masses of fibres, their shape indicating that they are replacing regional kyanite. In places, small remnant kyanite grains survive at the cores of such fibrolite masses.

Staurolite is comparatively rare in these rocks, and for the most part is confined to very small inclusions within garnet. However, Goodman (1993) recorded at [NO 3655 9180] small ragged grains of zincian staurolite, cored by gahnite (zinc spinel), which are remnants of originally larger porphyroblasts. The staurolite is believe to have survived the high metamorphic grades through the stabilising influence of zinc and the low availability of reactant quartz.

The semipelitic rocks west of the Glen Doll Fault show slight compositional and textural differences from those described above. For example, the plagioclase is oligoclase rather than andesine and may form locally prominent porphyroblasts such as those at Knapps of Fee around [NO 2416 7524]. Also the more pelitic parts commonly contain muscovite, sillimanite and, locally very abundant kyanite.

The psammites that characterise the upper part of the formation west of the Glen Doll Fault are around 100 m thick west and south-west of Foghorn, on the eastern scarp of Mackle Kilrannoch, and in upper Corrie Fee. They are locally gritty, and graded bedding is present [NO 2434 7491]. However, psammites are largely absent from parts of The Dounalt [NO 243 761] and the upper Mayar Burn section, implying that rapid local facies changes are present

There is a noticeable concentration of calcsilicate-rocks around the south summit of Craig Megen [NO 3175 8935] where they occur interbedded with gneissose semipelite and pelite, in zones up to 15 m wide. They comprise massive or, more rarely, foliated pale greenish rocks which may consist almost entirely of diopside (S82866), or a granoblastic intergrowth of plagioclase and diopside with subordinate amphibole and biotite, and accessory sphene and quartz. Carbonate-bearing rocks are comparatively rare in the Queen’s Hill Gneiss Formation, but brown-weathering impure limestone within a 4 m-thick calcsilicate-rock unit and, elsewhere, loose blocks of white crystalline limestone confirm the presence of such rocks at two localities in the Craig Megen area. A calcsilicate-rock exposed in Middle Grain [NO 355 869] contains up to 25 per cent calcite. It occurs in a fault-bounded block, so its stratigraphical position within the formation is unclear.

A 1.5 m-wide, thinly banded, impure pyritic limestone forms a minor channel [NO 2400 7473] for part of the Fee Burn. In thin section (S83930) calcsilicate-rock from this unit comprises quartz, plagioclase (generally calcic andesine, and zoned possibly with labradorite in central part), diopside, calcite and clinozoisite with minor ilmenite, ragged biotite and sphene lozenges. A thinner bed of calcsilicate-rock occurs some 10 m away in the main course of the burn. Both occurrences of calcareous material lie within highly migmatitic and partly pegmatoid gneissose semipelite. Calc-silicate pods and lenticles are common in the upper part of the formation, for example on the east flank of Meikle Kilrannoch [NO 2306 7708], in the upper part of Mayar Burn around [NO 2344 7360], and in the Knapps of Fee [NO 2418 7530]. Under the microscope (for example (S83931)), they consist of quartz, plagioclase (sodic andesine to labradorite), garnet and biotite with minor ilmenite (and related sphene), green hornblende and late- stage clinozoisite, chlorite and small muscovites. Sericitisation of plagioclase is widespread.

The vast majority of the metasediments from this formation have low magnetic susceptibilities (< 0.07 x 10−3 SI). A few exceptions have susceptibilities as high as 3 x 10−3 SI, and the local variability caused by these more magnetic units helps to pick out the main regional trend across the area.

Metabasic components

The Queen’s Hill Gneiss Formation contains a large volume of metamorphosed basic igneous rocks, particularly in the northern part of the outcrop. They consist mostly of concordant units of hornblende gneiss ranging in thickness from a few centimetres to several tens of metres, but on the north-west side of Glen Muick, within the confines of the Coyles of Muick shear zone there are significantly larger (up to 750 m wide) outcrops of sheared and amphibolitised metagabbro and serpentinite. The majority of these rocks are metamorphosed intrusive rocks and are described in more detail in Chapter 9.

Balnacraig Metabasite Member (BC)

The metabasic suite in the Queen’s Hill Gneiss Formation also includes some hornblendic rocks which are thought to have a volcanic protolith, although such rocks have not been previously recorded in the Crinan Group outside the district. These rocks, which have considerably more restricted occurrence than their intrusive counterparts, are here named the Balnacraig Metabasite Member. Lithologically they are similar to the Meall Dubh Metabasites Formation; however, they form a geochemically distinct suite (Goodman and Winchester, 1993).

The outcrop of the Balnacraig Metabasite Member is confined to a 2.5 x 1 km rectangle of ground immediately south of The House of Glenmuick [NO 373 945], which is delimited by the Ballater Granite to the north-east and elsewhere by metasedimentary gneisses or faulting. The member is exposed chiefly in the gullies near Garlot [NO 35 93] and around the ruined settlement of Balnacraig [NO 35 92]. It consists mainly of fine-grained and finely foliated hornblende schists, with interlayered calcsilicate-rock. All of the samples collected have low magnetic susceptibilities (2 to 6 x 10−3 SI) and the member therefore forms no distinctive feature on the magnetic maps. The magnetic characteristics of the Balnacraig metabasites contrast with the much more variable susceptibilities observed in metabasic horizons in other formations.

Close to the top of the Balnacraig Metabasite Member there is a layer of Queen’s Hill-type gneissose pelite [NO 364 936]. However, the upper margin of the member is taken at the last appearance of the highly distinctive interbanded hornblende schist and calcsilicate-rock. On the north-west side of the outcrop the metabasites are faulted against Queen’s Hill gneisses, so the nature of the rocks that originally underlay the Balnacraig Metabasite Member is indeterminate.

The calcsilicate-rock layers vary from a few millimetres to 0.3 m in thickness, and can be laterally persistent or pod-like. They are everywhere concordant with the foliation in the surrounding hornblende schist. This calcsilicate-rock is generally massive and differs from those described previously in consisting predominantly of fine-grained epidote, with a small amount of quartz, and scattered large porphyroblasts of diopside, but no free calcite. Piedmontite is present within some samples, apparently as a retrogressive phase. In places there are cross-cutting layers of a more feldspathic composition, although they may still contain some calc-silicate minerals. These were probably originally metasomatic veins, and may be strongly discordant, or smeared into parallelism with the amphibolite foliation. In places the podded appearance of the hornblende schists, with calc-silicate material between elongate pods, gives the rocks the appearance of deformed pillow lavas. However, there is a very strong planar fabric within the amphibolites and it seems unlikely that any such large-scale textural evidence for their origin would have survived the intense strain that the rocks have suffered.

The major and trace element geochemistry of the amphibolites, as determined by XRF analysis, is illustrated in (Figure 11), and full analyses can be found in Goodman et al. (1990). In common with other Dalradian metabasites (Graham, 1976b) the Balnacraig rocks are enriched in TiO2 and lie close to the boundary between the alkaline and tholeiitic fields when plotted on a TiO2 v Zr/P2O5 diagram. These are relatively immobile elements and are taken to reflect the original (premetamorphic) composition of the rocks. On an AFM diagram (Figure 11) the metabasites display the typical tholeiitic iron-enriched trend which again is unlikely to have been affected by metamorphism (Graham, 1976a). Compared with the Meall Dubh Metabasite Formation and the Farragon Beds these rocks are enriched in Nb (Goodman and Winchester, 1993). Several of the metabasites collected from the Balintober area [NO 36 94] show relatively low high field strength elements (for example, Ce, P, Zr, Hf, Sm, Ti, Y, Yb), and Th and Nb enhancement more characteristic of some volcanic arc rocks.

The Balnacraig metabasites are considered to have originally been formed as basic lavas of tholeiitic affinity, possibly submarine, and with thin layers of sediment deposited between eruptions. The absence of free calcite suggests that the original sediments were not calcareous. Certainly, there is enough calcium within the amphibolites for the present calc-silicate mineralogy to have derived by metasomatism of pelitic material.

Tayvallich Subgroup

Throughout much of the central and south-west Highlands the Tayvallich Subgroup consists of just one calcareous formation variously known as the Loch Tay or Tayvallich Limestone (Harris et al., 1994). The limestone has a remarkable lateral persistence and can be traced for over 200 km north-eastwards from Kintyre to Glen Isla in Angus. However, north-east of Glen Isla there is a marked increase in the amount of clastic material at this stratigraphical level so that, in the northern part of the Ballater district the main unit of carbonate rocks forms only a small proportion of the total thickness and is confined to the basal part of the subgroup. In consequence, north-east of the Glen Doll Fault the Tayvallich Subgroup can be divided into two formations, the Water of Tanar Limestone and the Tarfside Psammite formations. West of the fault the subgroup is represented by a single formation, the Dounalt Limestone Formation.

Water of Tanar Limestone Formation (bCT)

The ribbed calc-silicate and limestone rock of the Water of Tanar Limestone Formation directly overlies the Queen’s Hill Gneiss Formation and, as such, is correlated with the Loch Tay Limestone in the Pitlochry district. The junction between the two formations is sharp, and the Queen’s Hill metasedimentary rocks do not become calcareous toward the contact. Lying above the calcareous horizon, and everywhere associated with it, is a gneissose psammite or semipelite and a metabasic unit, both of which are also included within the formation. This association of rock types is distinctive and laterally continuous for several kilometres, and so is distinguished from the overlying Tarfside Psammite Formation, which also contains psammitic, metabasic and calcsilicate-rocks.

The metasedimentary rocks of the Water of Tanar Limestone Formation are traceable from the Pollagach Gap [NO 40 93] southwards to the F3 fold nose around Gowan Knowe [NO 33 81], where they pinch out. Their termination occurs at the same north-west-trending lineament as the more subtle variation in the metasedimentary rocks of the Queen’s Hill Gneiss Formation (p.37), which also marks a south-easterly change from psammite to semipelite in the overlying Tarfside Psammite Formation. In contrast, the metabasic unit (D1) continues to follow the junction between the Queen’s Hill Gneiss and Tarfside Psammite formations south-westwards to the margin of the Glen Doll Diorite at [NO 29 78].

The hornblende gneiss forming the metabasic unit has a distinctive magnetic signature which makes the formation readily traceable through areas of poor exposure. Bedrock samples from the eastern margin of the unit have susceptibilities as high as 10 x 10−3 SI, and the metabasite forms a prominent high-amplitude anomaly along the eastern boundary of the formation. Specimens from all other lithologies in this formation tend to have low magnetic susceptibilities (< 0.1 x 10−3 SI). The calcsilicate-rocks contain enough pyrrhotite to produce susceptibilities up to 0.15 x 10−3 SI, and on the magnetic profiles a low-amplitude anomaly marks the location of the carbonate horizon.

The calcareous rocks which lie at the base of the formation typically consist of a grey, crystalline, ribbed limestone and calcsilicate-rock, with some more psammitic bands. These are well exposed in the Water of Tanar [NO 37 88], where the stream runs approximately along strike, exploiting the comparative softness of these rocks. The content of free calcite in these rocks varies from 5 to 25 per cent. Epidote group minerals, particularly clinozoisite and pistacite, are common as small rounded grains, normally associated with larger and more idioblastic grains of diopside. Quartz usually forms 25 per cent or more of the rock, together with plagioclase (anorthite or bytownite), microcline, hornblende, and pale phlogopitic biotite. Small grossular garnets are common, and may show considerable anisotropy. Sphene may be so abundant as to be considered a major mineral, and apatite, zircon and opaque minerals form the accessories. The ribbed limestone may contain 2 to 3 per cent disseminated sulphide, and although most of this is pyrite and pyrrhotite, some chalcopyrite and bornite are also present. The calc-silicate minerals wollastonite, scapolite and idocrase are present in small amounts in some samples, but are abundant in those calcsilicate-rocks which have a thermal metamorphic overprint (Chapter 13).

There tends to be little foliation within the calcsilicate-rocks, and they have a granoblastic texture, with a close mosaic of generally equigranular minerals. Commonly there is a compositional banding, transposed into the direction of the main regional foliation, which reflects the variation in the amount of calcite present and results in a ribbed outcrop.

The psammite and semipelite that overlie the limestone are typically calcareous in part. These rocks are generally migmatised, and associated (for example in the Water of Mark [NO 3542 8371]) with the calcareous rocks there is commonly a coarse-grained white gneiss, which is probably a leucogranitic segregation. It contains quartz, muscovite and K-feldspar; the last is absent from leucosomes derived from non-calcareous metasediments. Where the calcsilicate-rocks are veined, the vein material is usually quartz, and there does not seem to have been appreciable remobilisation of the calcite. The associated metabasic unit is a fine, garnetiferous hornblende schist.

Dounalt Limestone Formation (bLT, QT, QLT and QCT)

This dominantly calcareous formation, which ranges from less than 10 m up to about 100 m thick, has a stratigraphical significance disproportionate to its thickness. It forms a mappable unit which defines both the stratigraphy and structure. Barrow (1893) recorded some limestone localities on his field maps but did not assign them stratigraphical significance. Lithologically the unit has features intermediate between the Loch Tay Limestone Formation to the south-west and the Water of Tanar Limestone and Tarfside Psammite formations to the north-east.

The formation is exposed mainly in cliff and stream sections, for example on Erne Craigs in Corrie Fee [NO 2461 7548], in a small steep gully at the south-east end of The Dounalt [NO 2476 7602], in the White Water [NO 2344 7758] and at localities upstream as far as [NO 2267 7836], in the Burn of Dounalt [NO 2411 7655] and the tributary to the south at [NO 2415 7634], in the Burn of Fialzioch [NO 2349 7732], and in the Fee Burn [NO 2442 7495]. Minor exposures of limestone and pyritic quartzite have been found along strike to the south-south-west in White Glen [NO 2294 7234]. Farther north thinly banded fawn/turquoise calc-silicate units are exposed at the south-east end of Crow Craigies [NO 2266 7950]. Thereafter, the outcrop of the formation is believed to pass eastwards through largely unexposed ground for about 2 km before it is truncated by the southern edge of the Lochnagar Granite. Possible indications of a further easterly extension are to be found in quartzose psammite and calcsilicate-rock screens within the Juanjorge Diorite, which are well seen in the bed of the River South Esk between [NO 2613 7927] and [NO 2627 7920].

The Dounalt Limestone Formation consists of thin- to medium-banded crystalline limestones and calcsilicate-rocks, commonly with siliceous and semipelitic interbeds, calcareous psammites, semipelites, quartzites and rare gritty quartzose psammite, for example at [NO 2339 7764]. Pyrite is notably abundant in these lithologies. The thickness variations of the limestone partly reflect original facies variations, but also result from local structural effects. For instance, in cliff sections along the Burn of Dounalt and on Erne Craigs, thinly banded limestones with minor calcsilicate-rock, psammite and semipelite interbeds are repeated by minor folding such that they have structural thicknesses of 35 m and 20 m respectively. Similarly, in the White Water [NO 2335 7761] thinly banded limestones and calcareous quartzose and micaceous psammites are tightly folded (F3 + F2) as shown in (Plate 8). In the Fee Burn section, the formation is represented by two beds, 3 m and 1 m thick, of blue-white to pale grey crystalline limestone, with calcareous semipelite interbeds, separated by about 10 m of quartzose psammite and interbanded psammite and migmatitic semipelite. Farther to the south-south-west in White Glen only small exposures and abundant float of limestone, pyrite-rich calcsilicate-rock and quartzite are seen; the limestone debris weathers to a pale fawn to cream colour, and is strongly rodded and thinly bedded.

Banded calcsilicate-rocks are common within the limestones, but at the south-east end of Cow Craigies adjacent to Jock’s Road, the entire formation is represented by pale fawn, turquoise or pink, thinly banded, locally pyritic calcsilicate-rock units. In thin section (S92791) diopside and large (up to 5 mm) poikiloblastic grossular garnet are seen to lie in a matrix of plagioclase, scapolite, tremolite and minor quartz, calcite, scattered sphene, and late stage muscovite.

Typical assemblages for these impure limestones and calcsilicate-rocks are:

Microcline occurs in a calcsilicate-rock (S82917) from Erne Craigs [NO 2472 7548]. Specimens from the Dounalt Burn (S82959) contain abundant idocrase, possibly a product of contact metamorphism by the underlying granodiorite body or the related overlying microgranodiorite/microgranite dykes.

Psammite units occur within the formation both above and below the limestone itself. In the White Water section between[NO 2344 7758] and [NO 2273 7825] quartzose to micaceous psammite, and minor semipelite are interbedded with limestone and calcsilicate-rocks. The resultant thinly banded lithology commonly shows tight minor folding for example at [NO 2291 7789].

South-west of Craig Rennet [NO 2506 7575] quartzose psammite is interbedded with gneissose semipelite and micaceous psammite for up to 20 m stratigraphically below the limestone. North-west of Craig Rennet, psammite generally forms only a minor component in the dominantly gneissose semipelite below the limestone. However, a thick lenticular unit of gritty feldspathic psammite does occur immediately to the north-west of Craig Rennet; it is locally calcareous and contains abundant thin bands of calcsilicate-rock, and weakly banded impure crystalline limestone units up to 1 m thick. The psammite has been thickened by early folding but was at least 50 m thick prior to deformation. Minor migmatitic gneissose semipelite occurs towards the base of the gritty psammite, close to the limestone. Dominantly psammitic units, locally gritty and with calc-silicate lenses are also found in the Burn of Dounalt and adjacent stream sections, at the base of Craig Maud [NO 240 770], and in the White Water below the Lunkard. At around 400 m south-west from Cairn Damff, gneissose psammite that is locally gritty is exposed over a wide area. In thin section (S80864) recrystallised quartz clasts occur in a foliated quartz-plagioclase-biotite matrix with accessory apatite, ilmenite and monazite. In parts this gritty lithology has a granitoid appearance and consists of quartz-muscovite-biotite-plagioclase (S80863). On the south slope of Cairn Damff [NO 2397 7759] near-horizontal gritty psammite and minor semipelite show graded bedding implying that the beds are right way up. The psammites are similar to those adjacent to the limestone and are taken to be an inlier of the Dounalt Limestone Formation within the Southern Highland Group rocks.

In some localities, for example in the Fee Burn section in Corrie Fee [NO 2440 7494]) a hard, fine- to medium-grained, massive amphibolite, some 10 m thick, directly underlies the limestone unit. This amphibolite is comparable to that found in the Water of Tanar Limestone Formation to the north-east, although magnetic susceptibility values are significantly lower, being typically 0.4 x 10−3 SI. Amphibolite (metadolerite and metabasalt) sheets are abundant elsewhere within the Dounalt Limestone Formation.

Tarfside Psammite Formation (SQT and bQamT)

East of the Glen Doll Fault the Water of Tanar Limestone Formation is succeeded by the Tarfside Psammite Formation (Figure 7). This formation is characterised by an along-strike change from predominantly psammites and quartzites in the north, the Glen Tanar Quartzite Member (bQamT), to a preponderance of semipelites in the south, the Cald Burn Gneiss Member (SQT). The Glen Tanar Quartzite Member comprises some of the rocks described by Harte (1979) as the ‘Tarfside Group’ which he considered to include equivalents of both the Loch Tay Limestone and Ben Lui Schist. Read (1928), on the other hand, considered these rocks to be equivalent to the Pitlochry Schist.

Glen Tanar Quartzite Member (bQamT)

The Glen Tanar Quartzite Member can be traced for about 17 km, from the eastern edge of the district between [NO 423 940] and [NO 424 954] to the southern slopes of Easter Balloch [NO 348 795], where it is replaced along strike by the Cald Burn Gneiss Member. The outcrop width in the north is limited by the intrusion of the Mount Battock Granite, but south of the Water of Mark it exceeds 6 km. The overall strike of the formation is north-north-east, but from Craig Damff [NO 380 790] to Drumhilt [NO 353 805] the contact with the overlying Longshank Gneiss Formation runs west-north-west for about 3 km, as it passes between two F3 fold closures. No such flexure is evident in the lower contact of the formation with the result that outcrop width is reduced to 1.5 km at Drumhilt. This rapid thinning coincides with the passage to the Cald Burn Gneiss Member.

The member comprises interbedded psammites, quartzites, calcsilicate- and amphibolitic rocks. The psammitic rocks show a range in their degree of maturity and silica content. They contain varying proportions of biotite, present in plates of moderate size, which, along with elongation of quartz and feldspar, define the fabric of the rock. The feldspar is predominantly andesine, but some K-feldspar may also be present. In some psammites, there are small, idioblastic almandine garnets, and opaque minerals, together with accessory apatite and zircon. Local thin pelite beds within the Glen Tanar Quartzite Member are similar to the pelites of the Queen’s Hill Gneiss Formation, again containing quartz, andesine, K-feldspar, dark biotite, sillimanite, almandine and accessories (opaque minerals, apatite, zircon, sphene). In the north, all lithologies of the Glen Tanar Quartzite Member tend to be more pervasively calcareous, with the psammites containing grossular garnet, for example near Knockie Branar [NO 4037 9272]. A calcareous psammite (S79440) collected from the southern slopes of Hunt Hill [NO 3823 7993] contains quartz, carbonate, microcline, plagioclase, diopside, sphene and actinolite in decreasing order of abundance.

Calcsilicate-rocks occur throughout the member, typically in the form of lenticular bodies up to 10 m in thickness, and commonly exhibit a prominent brown-weathering surface. A range of mineral assemblages is present and comprises variable proportions of quartz, plagioclase, microcline, carbonate, tremolite, actinolite, hornblende, diopside, sphene, clinozoisite, phlogopite, biotite and grossular garnet. Mineral proportions vary widely both between and within outcrops, the rocks ranging from grey impure limestone containing at least 70 per cent of free carbonate, to pale green rocks dominated by calc-silicate minerals. Some are quartzose with significant amounts of microcline. Many possess a millimetre- to centimetre-scale lithological lamination, resulting mostly from alternating quartzofeldspathic and amphibole-rich layers.

The member also contains abundant units of hornblende gneiss which are generally coarse grained and agmatitic, and concordant with the adjacent metasedimentary rocks. In places they are present as discrete pod-like developments, rather than laterally continuous sills, as in the Queen’s Hill Gneiss Formation. However, they are considered to be similar in origin to the metabasic rocks of the Queen’s Hill Gneiss and Water of Tanar Limestone formations.

Metasedimentary rocks of the Glen Tanar Quartzite Member have low magnetic susceptibilities of less than 0.4 x 10−3 SI, with the exception of one specimen of calcsilicate-rock which gives a value of 8 x 10−3 SI. The hornblende gneisses give values around 10 x 10−3 SI, and these, along with some calc-silicates, produce the distinctive low-amplitude MAGGRO anomalies that characterise this member. The amphibolitic mineralogy of some calcsilicate-rocks together with the quartzose nature of some hornblende schists may indicate a tuffaceous origin for some rocks in the member.

Cald Burn Gneiss Member (SQT)

The most northerly occurrence of the Cald Burn Gneiss Member is on the north-east flank of Muckle Cairn [NO 353 825] where it forms a 300 m-wide unit between the Water of Tanar magnetic hornblende schist and the Glen Tanar Quartzite Member. Southwards, it increases in thickness progressively at the expense of the Glen Tanar Quartzite Member, so that between Easter and Wester Balloch [NO 34 79] it constitutes the whole of the Tarfside Psammite Formation. The outcrop can be traced for a further 5 to 6 km to the south-west, to where it is truncated by the Glen Doll Diorite, giving a total strike length of 8 to 9 km. The Cald Burn Gneiss Member is predominantly semipelitic with minor pelitic and psammitic units; no quartzites and only minor calcsilicate-rocks have been recorded.

The gneissose semipelite is composed of an inequigranular aggregate of quartz, plagioclase, red or brown biotite, muscovite and garnet. The range in grain size is largely attributable to the sporadic development of some large grains of plagioclase and more rarely of quartz. Overall, biotite is the dominant mica, although locally muscovite forms larger crystals and is more common. In places muscovite is concentrated in narrow but persistent foliation-parallel laminae. The foliation is largely defined by common orientation of the micas, and in places by plagioclase and quartz.

Pelites within the Cald Burn Gneiss Member differ from those in formations described above. They contain quartz, andesine and K-feldspar, but have a rather greenish biotite, and muscovite in stable association with the other minerals. Such differences are thought to be the product of differing oxidation states in the original sediments (Chinner, 1960). There may be a small amount of sillimanite; garnets are common and are usually small, unzoned, idioblastic and inclusion-free. The composition of the garnets is typically Alm77Py12Gr6Sp5. Opaque minerals, apatite and zircon are the common accessory minerals.

Measurement of magnetic susceptibilities from specimens of the Cald Burn Gneiss member are all distinctively low, accounting for the featureless area on magnetic maps occupied by this member. This contrast serves to distinguish the two members of this formation, even at the low-amplitude variation involved.

Chapter 8 Dalradian Southern Highland Group

On gross lithological terms the Southern Highland Group succession in the Ballater district is divisible into two formations (Figure 7). This subdivision is broadly comparable to that established by Harte (1979) in the Effock–Lethnot Group of the Tarfside succession (in the Aboyne district) where the Glen Effock Schist is overlain by the Glen Lethnot Grits. However, marked changes are present in the lithology across the Glen Doll Fault (Figure 12). East of the fault the basal unit of the Southern Highland Group, the Longshank Gneiss Formation, is a thick sequence of semipelites with psammites and pelites. It is notable for its high magnetite content and absence of metabasic rocks. The overlying Rottal Schist Formation consists of schistose semipelite and psammite in upper Glen Clova, changing laterally to the east to a greywacke-dominated sequence, which also includes Green Beds. The thick pile of amphibolites within the formation west of Glen Clova may represent a basic volcanic sequence, but it is more probable that the amphibolites were largely basic intrusions into the sedimentary sequence.

West of the Glen Doll Fault, the basal unit of the Southern Highland Group is known as the Pitlochry Schist Formation. It is a relatively thin pelite- and semipelitedominated unit; unlike the Longshank Gneiss Formation it is only weakly magnetic. In contrast to the succession to the east, amphibolites, which are interpreted as mainly high-level sills, are notably abundant in the whole Southern Highland Group sequence west of the fault. The Rottal Schist Formation in this area is more variable than to the south-east. It contains more semipelite, and shows variably developed volcanic and volcaniclastic Green Bed lithologies in the lower part of the sequence. Gritty units that show graded bedding are recorded in Mayar and White glens just above Green Bed lithologies.

Longshank Gneiss Formation (SQsH)

The Longshank Gneiss Formation has a south-westerly tapering wedge-shaped outcrop which ranges in width from 6 km around Loch Lee to less than 2 km where it is truncated by the Glen Doll Fault. The shape of the outcrop is largely the result of F3 folds; this is most apparent in a series of westerly verging steps in the upper boundary and to a lesser extent in the lower. Rocks of the formation are particularly well exposed on the extensive crags that cap the upper slopes of the deeply incised valleys of the waters of Mark, Lee and Unich. There are also excellent sections in the walls of the high-level corries on the north-east side of Glen Clova. However, the most extensive exposures are to be found on the south-west side of Glen Clova where the formation can be traced almost continuously along strike for over 6 km, and in places there is a near-complete vertical section of up to 400 m.

The Longshank Gneiss Formation consists principally of gneissose semipelite, pelite and minor psammite and is entirely devoid of interbanded metabasic rocks. The rocks contain abundant magnetite, and measured magnetic susceptibilities range from 0.2 to 25 X 10−3 SI with about 30 per cent of the unit giving values of between 2 and 20 x 10−3 SI. The enhanced magnetic susceptibilities occur in all three principal lithologies, but the high values are confined to certain units so the formation as a whole has a spiky magnetic profile. Magnetite is accompanied in the north and east by tourmaline and, since both minerals are obviously of detrital origin, it follows that the magnetic signature is a primary feature and can therefore be regarded as a valid strati-graphical marker.

In the north, the lithological contrast between the rocks of the Glen Tanar Quartzite Member and the Longshank Gneiss Formation is such that the boundary can be followed readily in well exposed ground. No such contrast exists between the Longshank Gneiss Formation and the Cald Burn Gneiss Member; consequently in the south and in areas of poor exposure the contact is based entirely on the magnetic contrast between the two units. The upper boundary of the formation is more readily discernible in the field because of the lithological contrast between it and the overlying rocks, the Rottal Schist Formation, or with hornblende schist and gneiss to the south-west of Loch Brandy. On the south-west side of Glen Clova the boundary can be followed for over 7 km, initially up the crags above Atton, thence round the F3 Corlowie Fold and across the summit of Bassies to the Sneck of Farchal [NO 282 736], where it is truncated by the Farchal Fault. It reappears on the opposite side of the fault roughly 300 m to the south from where it can be traced along the southern slopes of Driesh and Little Driesh [NO 254 726], beyond which it swings south-westward into the valley of the Burn of Kilbo.

There is a subtle lateral facies variation in the Longshank Gneiss Formation, along the same line as noted in units lower in the succession (Figure 8). In the north and east, for example around Drumhilt [NO 35 80], the rocks are mixed semipelites and pelites, with common psammite beds, whereas in the south and west, for example on Cathelle Houses [NO 31 76], they are more monotonously semipelitic to pelitic, with some micaceous psammite, but only rare psammite. The volume of interbedded psammite increases again on the south-west side of Glen Clova although the succession is still dominated by gneissose semipelite and pelite. However, the predominance of psammite on the summit plateau of Driesh [NO 27 73] is such that a separate unit, the Driesh Psammite Member (QSSH), has been identified. The member consists predominantly of psammite to micaceous psammite with subordinate beds of gneissose pelite and semipelite, and can be traced over Little Driesh [NO 26 72] to the Glen Doll Fault. Despite being the oldest part of the succession between the Glen Doll and Farchal faults, and lithologically similar to the Tarfside Psammite Formation, the member is still considered to be a part of the Longshank Gneiss Formation, mainly on the evidence of its high magnetic susceptibilities. The fact that on Little

Driesh the upper boundary of the member comes close to the top of the formation would seem to confirm that this represents a facies variation within the formation.

The Longshank Gneiss Formation is dominated by muscovite-rich gneissose semipelite which has a well-developed foliation defined by alternating quartzofeldspathic and micaceous layers, and parallel alignment within the micas. The leucosomes comprise quartz and andesine in varying proportions with minor muscovite and biotite. A few samples additionally contain K-feldspar.

The melanosomes consist essentially of muscovite, biotite, garnet and magnetite. Muscovite in many cases is dominant over biotite, forming large plates, mostly foliation parallel but also cross-cutting. Biotite is mostly pleochroic from yellow to dark brown, but in parts is distinctly greenish in rocks with elevated magnetic susceptibility. The garnets are generally euhedral and tend to be concentrated in biotite-rich domains, particularly those containing green biotite. In most rocks they are small and inclusion free, but in a number of samples they are somewhat larger and enclose quartz and more rarely magnetite and zircon. Kyanite, as small crystals entirely enclosed within muscovite, is confined to the upper part of the formation. Minor amounts of fibrous sillimanite, with similar occurrence have more widespread distribution.

Pelitic beds within the succession have a number of features which, in addition to the higher proportion of micas, distinguish them from the semipelites. In particular, they contain staurolite (albeit rarely) and tourmaline together with abundant kyanite, garnet and magnetite. Also, biotite is mostly reddish brown, although deep brown and greenish varieties also occur. Garnet is ubiquitous and much coarser than in the semipelites, occurring as porphyroblasts up to 5 mm in diameter. Inclusions in garnet are common; these consist mostly of lobate quartz, together with muscovite and biotite which are generally parallel with those external to the garnet; zircon and staurolite inclusions are rarer.

Kyanite is a prominent component of pelites within the upper part of the formation, but was not recorded in the higher grade rocks to the north-west. Crystals are mostly less than 5 mm across, but are up to 30 mm in Corrie Kilbo [NO 2689 7458]. In places, for example at Bassies [NO 2957 7358], the kyanites are obviously aligned along the principal lineation in the rock. This preferred orientation is very evident in thin section mainly because the porphyroblasts generally display a prismatic habit. Other forms are also present. For instance in thin section (S81032) the habit ranges from well-cleaved and inclusion-free elongate laths, through distinctly more stumpy and characteristically poikiloblastic grains full of rounded quartz, aligned biotite and magnetite trails, to much smaller uncleaved anhedral crystals that form a granoblastic polygonal intergrowth with quartz and biotite. The poikiloblastic variety strongly resembles some of the thermal andalusite seen in rocks of the Creag nam Ban Formation (Chapter 6). The fact that all three varieties occur within the same rock would seem to suggest that the difference relates to the nature of the mineral on which it nucleated. For instance in thin section (S79595), kyanite appears to have grown at the expense of both staurolite and muscovite which, if taken to completion would have resulted in markedly different forms. Fibrolite is present to a limited extent in the kyanitebearing rocks, forming thin and discontinuous anastomosing zones generally within muscovite-rich domains. In places it appears to be nucleating on staurolite or garnet, but is seldom seen in contact with kyanite. Accessory minerals in the pelites are restricted to zircon and, more locally, tourmaline, with secondary chlorite.

Magnetite is present in more than accessory amounts and in places is accompanied by hematite. These iron-rich gneisses contain bright green (as opposed to reddish brown) biotite and far more muscovite than the hematite-free gneisses.

Pitlochry Schist Formation (SSH, PSH, and PSSH)

This formation includes the major occurrences of semipelite and pelite in the Glen Doll–upper Glen Clova area. A band of locally tightly folded migmatitic garnetiferous semipelite and micaceous psammite occurs immediately east of the attenuated limestone–quartzite Dounalt Limestone Formation in upper White Glen, and this can be traced northward to Mayar Glen and Corrie Fee. In south Corrie Fee, gneissose semipelite with quartzose and micaceous psammite layers are generally permeated by pegmatite segregations and in parts, [for example around [NO 2505 7468] are massive, coarse grained and granitoid. Pegmatite patches and relict amphibolite, quartzose psammite and pelite are abundant. The pelite fragments are small, ragged, and ill-defined. In thin section (S82885) the granitoid rock is coarse grained (1 to 5 mm) and consists of equant quartz grains, plagioclase (oligoclase/andesine) laths, and large muscovite, biotite and K-feldspar grains, with minor small rounded garnets and accessory monazite and zircon. The plagioclase is zoned in places, and the muscovite postdates and in part replaces the biotite. The rock probably formed by partial melting of semipelite.

In the north-west part of the Glen Doll–upper Glen Clova area, gneissose pelitite is well developed immediately overlying the Dounalt Limestone Formation. Kyanite is common within this lithology, typically forming small blue-grey laths about 5 mm long. In places it occurs as clusters of interlocking silver-grey to pale blue laths, individually up to 12 mm long; for example west of Dounalt [NO 2361 7647] where it occurs in massive, coarsely foliated, schistose pelite. In thin section such pelites (S80889), (S80879) consist of large corroded and altered kyanites (generally 2 to 4 mm long), large poikiloblastic low-andesine plates, subsidiary quartz, clusters of biotite and muscovite, small slightly altered garnets, and swirls and felts of fibrolitic sillimanite. Magnetite is common particularly as small grains and rods associated with biotite margins. The corroded kyanites commonly lie in a fine radiating aggregate of sericite/muscovite. Gneissose pelite and semipelite crop out widely in the Cairn Lunkard–Crow Craigies–Loch Esk area around [NO 235 790], where the Pitlochry Schist Formation is duplicated in the hinges of complex F2 + F3 fold patterns. Exposure is poor in part of this area and the bulk of the unit has been strongly hornfelsed.

Around the south-west summit of the Craigs of Loch Esk [NO 2376 7864] massive, granitoid, crudely foliated, oligoclase-biotite-rich gneissose semipelite is well exposed. Here and to the south-west, minor psammite bands show the original compositional banding. Pink tabular thermal andalusites up to 10 mm long are seen in sillimanite- and kyanite-bearing, biotitic, gneissose pelite and semipelite around [NO 2367 7860]. Cordierite is also present both here and on the north-west flank of Cairn Lunkard, and farther to the north-west around [NO 2309 7905]. At this last locality blocky, purplish tinged dark grey, foliated semipelite and pelite contain pearly white radiating aggregates of fibrolite and in parts small purple garnets. In thin section (S80907) fresh cordierite and andalusite porphyroblasts overgrow the foliated regional assemblage of plagioclase-biotitefibrolite-quartz-muscovite.

In the Craig Mellon–Cairn Broadlands–Dog Hillock area to the north-west of the Glen Doll Diorite, a thick sequence of regularly foliated biotite-rich gneissose semipelite and minor micaceous psammite is seen. More psammite-dominated parts are seen locally, for example adjacent to the diorite on the steep south-east slope of Cairn Broadlands. Minor calc-silicate lenses are present near the margins of the semipelite unit, notably in the area south-south-west of Craig Mellon. Pegmatite segregations are very abundant in the semipelite, typically as veins and pods subparallel to the foliation. Oligoclase porphyroblasts are locally common and in parts the semipelite takes on a granitoid appearance. The outcrop pattern suggests that the Pitlochry Schist Formation in this area is tightly folded, probably with interference structures, but in the field, the later foliation dominates and only rarely are minor folds recorded. In thin section (S83942) the semipelite consists of quartz-oligoclasebiotite layers with lenticles of slightly altered cordierite and small irregular andalusite porphyroblasts separated by coarse-grained quartz-oligoclase segregation veins. The presence of cordierite and andalusite in the semipelite some 550 m from the diorite contact attests to the extent of contact metamorphism which apparently affects the entire outcrop of the Pitlochry Schist Formation in this area.

Rottal Schist Formation (QgPRS, SQRS, QSRS and V)

These are the youngest of the Dalradian rocks exposed in the Ballater district and, as a result of the regional southeasterly sense of younging, occupy the southern extremity of the district. The formation is made up principally of metasedimentary rocks and Green Beds, but also includes hornblende schist and amphibolite which locally dominate the outcrop. These metabasic rocks are predominantly intrusive sheets and as such are described in Chapter 9. However, as these rocks are in places closely associated with Green Beds — regarded as volcaniclastic metasedimentary rocks — and elsewhere show a fine-scale interbanding with metasedimentary rocks, it is probable that they include metabasalts.

Metasedimentary rocks

The metasedimentary member constitutes the whole of the formation in the south-east corner of the sheet, but with the abrupt appearance of Green Beds on the west side of White Hill [NO 38 73] to [NO 38 74] its outcrop bifurcates. The northern prong can be followed westwards for a further 6 km to the Loch Brandy area. South-west of this, the outcrop of the metasedimentary rocks is reduced to discontinuous strips within hornblende schist and amphibolite, seldom more than 100 m wide, although cumulatively they occupy up to 40 per cent of the outcrop in the Carn Inks area [NO 30 72] to [NO 31 72]. Under the influence of F3 folds, the regional strike is step-like, alternating between north-east and north-west. West of the Glen Doll Fault, the outcrop width is around 1 km to the south-west of the Glen Doll Diorite. North and west of the intrusion, the metasedimentary rocks underlie much of the Glen Doll–upper Glen Clova area, although only the lower part of the formation in represented.

The formation consists of micaceous psammite interbedded with pelite and semipelite, although local compositional and mineralogical differences are apparent across the Glen Doll Fault. Thus in the south-east of the district, the succession consists mainly of micaceous psammite and pelite with only minor semipelite, whereas to the west of the fault it consists mainly of micaceous psammite and semipelite with pelite as only a minor constituent. In addition the western succession can be divided into an upper part, confined to the ground south-west of the Glen Doll Diorite, in which semipelite is more common than psammite, and a lower part where the reverse is true. A good section through the lower part is present on the crags on the north-east side of Glen Doll, for example on Craig Damff [NO 24 77]. The lowest parts of the formation are also well exposed on the steep lower crags of Craig Rennet. The interbanding of the lithologies generally ranges from centimetre- to metre-scale, but in the south-east may be up to tens of metres.

The psammites range from quartzose to highly micaceous, and from coarse and distinctly gritty to finely laminated. In the south-east, the gritty rocks are characterised by clasts, up to 6 mm in diameter, of white feldspar and colourless quartz in varying proportions. Ill-defined graded bedding was noted at a number of localities for example [NO 3294 7175]. The turbiditic origin of the metasedimentary rocks is readily apparent where these graded gritty psammite units are present. A pervasive spaced cleavage is defined by markedly lenticular and anastomosing zones of biotite and muscovite and a fine-grained mosaic of quartz-plagioclase, and clasts may be flattened along the cleavage. In thin section a relict clastic texture is also apparent in the more micaceous psammites, despite extensive recrystallisation.

West of the fault, partly graded gritty beds are rare, but good examples are seen in Mayar Glen [NO 2407 7313] and in White Glen [NO 2387 7207] and [NO 2372 7206]. At the former locality white to fawn, gritty, quartzose psammite grades from a coarse gritty base to a thin semipelite–pelite top. The beds dip moderately steeply to the west with a pronounced spaced cleavage (composite S1/2) dipping at a slightly shallower angle, also to the west. The beds young eastwards and hence are upward facing on S1/2 and inverted. In White Glen, reversals of facing are seen in blocky, similarly graded psammites, implying the presence of D2 or D3 tight folds. The psammites locally become gneissose where possible fold hinges occur.

The mineralogy of these rocks is quartz, K-feldspar, plagioclase, muscovite and biotite. Quartz is always the dominant mineral, and there is a variation in the relative proportions of the other minerals, particularly the micas. For instance, in thin section (S79317), muscovite is smaller but far more widely distributed than biotite which tends to be concentrated along with garnet, tourmaline and zircon in rather discontinuous zones, thought to be relict heavy mineral bands. In contrast, biotite is the dominant mica in thin section (S79298), a rock characterised by a relative abundance of garnets. A quartzose psammite (S80911) from west of the fault showed certain mineralogical differences, most notably mid-andesine plagioclase instead of K-feldspar and moderately abundant quantities of accessory ilmenite and magnetite. Minor garnet in this rock is heavily corroded. On Creag Damff [NO 24 77] the rocks range in composition from micaceous psammite to semipelite; banding is poorly developed and lenticular with common pegmatite pods and segregations. The psammite units, locally quartzose, are more coherently banded and commonly define tight early (F1 or F2) minor folds. Interference structures are seen [NO 2452 7731].

The semipelite is generally a dark grey and cream, lenticularly striped, coarsely foliated, biotite-rich gneiss. Pegmatite segregations are common and where semipelite lies adjacent to massive amphibolite there is commonly a zone of pegmatite 1 to 3 m wide at the contact, and particularly abundant veins and pods in the adjacent semipelite, for example on Craig Rennet [NO 2527 7572]. Oligoclase porphyroblasts are not common in semipelitic units within the Rottal Schist Formation, but are well seen locally in its basal parts on the southern crags of Craig Rennet and in parts of Corrie Fee. The Pitlochry Schist Formation is absent here, probably as a result of original facies variation, and the Rottal Schist Formation directly overlies the thin Dounalt Limestone Formation.

The semipelite found south-south-west of Mayar and adjacent to the Glen Doll Fault on the Shanks of Drumfollow and Drumwhallo [NO 24 72] to [NO 25 72] was portrayed on the previous 1-inch geological map (Sheet 65 Balmoral) as oligoclase-biotite gneiss. This survey has shown this to be incorrect; on the west flank of Cairn Dye around [NO 2466 7252] coarsely garnetiferous muscovitebiotite schist with thin beds of calcareous quartzose psammite and calc-silicate lenticles is exposed. Although quartz veins are common in this schist, it lacks the typical pegmatite pods and veins. Farther north-east, the semipelite becomes migmatitic on the Shanks of Drumwhallo and Drumfollow and on this latter ridge, some 400 m from the edge of the Glen Doll Diorite, it becomes massive with abundant pegmatitic segregations. Within 150 m of the diorite contact the semipelite is partly granitoid in nature. Such a rock (S83944) consists of a coarse-grained mosaic of unstrained quartz and plagioclase (An30), in part zoned, with abundant biotite and smaller (less than 1 mm across) anhedral, clear to patchy K-feldspar crystals and minor magnetite.

The pelitic rocks are dominated by a felt of commonly aligned muscovites enclosing garnet (up to 4 mm), biotite, quartz, plagioclase, staurolite and kyanite. Biotite is typically reddish brown and exhibits several habits, ranging from large equant porphyroblasts with ragged outline to small grains within the muscovite felt. Tourmaline is generally present at accessory level, but is a significant constituent of thin section (S79312) where it forms poikiloblastic grains up to 1.7 mm long. Hornblende is present in pelite and psammite on Craig Duff [NO 3279 7200]. Since both rocks are marginal to hornblende schist this may be further evidence that some of these metabasic rocks have a basaltic protolith.

The pelites west of the Glen Doll fault show less evidence of pegmatite segregation than the semipelite units. In the field they are typically fine- to medium-grained, hard, poorly banded, gneissose rocks with kyanite (relict) laths or muscovite after kyanite visible in hand specimen. As they are generally hornfelsed, the rocks have a glassy appearance and a characteristic dark purplish grey colour, and commonly comprise the assemblage cordierite-sillimanite-quartz-plagioclase-biotite-andalusite-ilmenite.

Calcsilicate-pods and lenticles, identifiable in the field by their pitted weathering surface are not uncommon within the psammites of the south-east succession, but in the west are more sparse although more common near the base of the formation. They commonly contain the assemblage quartz, plagioclase (altered), garnet, hornblende and sphene. Limestone was recorded [NO 32 71] only on the south side of Glen Clova, where it occurs as a single unit, up to 2 m thick, which can be traced along strike by limited exposure, float and vegetation changes for about 350 m. At its northernmost exposure [NO 3288 7189] the unit comprises a 1 to 2 m-thick bed of pale bluish grey limestone, dominated (S79311) by granoblastic calcite, but also including quartz, plagioclase, muscovite, biotite and hematite. This rock is also unusually rich in tourmaline and is interbedded with the tourmaline-bearing pelite described above. The associated psammite–pelite assemblage also includes two concordant units of actinolite schist, a rock type not seen elsewhere in the succession. Approximately 100 m to the south-west, the limestone is reduced to beds no more than 1 m thick within the tourmaline pelite. At the southernmost exposure it consists of a series of lenses seldom exceeding 0.5 m in thickness within highly contorted garnet pelite. The rock (S79328) is somewhat less pure than at the northern end with conspicuous biotite-iron oxide partings, garnet porphyroblasts and a much higher quartz-plagioclase content.

The metasedimentary strips within the thick amphibolite sequence in the south-east consist largely of psammite and pelite, with less common semipelite, although the latter occurs more commonly in the western part of the outcrop, most notably on Hill of Strone [NO 28 72] and on the south side of Driesh. The rocks are essentially similar to those elsewhere in the south-east, although gritty psammites are relatively uncommon. Also, because of increasing metamorphic grade, pelites in the west (for example (S80289)) contain fibrolite in addition to garnet, staurolite and kyanite.

Green Beds (V)

The Green Beds of the Southern Highland Group are generally accepted (for example Roberts, 1966; Van de Kamp, 1970) to be metamorphosed mafic volcaniclastic sedimentary rocks. Their main outcrop in the south of the district can be traced along strike for nearly 8 km, although much of the ground is poorly exposed. The outcrop is no more than 100 m wide in the south-west, but it gradually thickens eastwards to almost 1 km around the Burn of Heughs. The sudden termination of the Green Beds outcrop in this area was attributed by Harte (1979) to the effects of the Glen Mark Slide, but now it can be seen to coincide closely to the south-easterly extension of the zone across which the Queen's Hill Gneiss, Water of Tanar Limestone and Tarfside Psammite formations show rapid changes in lithology. Green Beds are well developed at the base of the Rottal Schist Formation, most notably in the Loch Brandy area, on the crags south-west of Atton [NO 3065 7336] to [NO 3059 7307] and on the south side of Driesh [NO 2787 7315] and [NO 2645 7266]. Isolated occurrences are also present within the thick sequence of metabasic rocks on Cairn Inks, for example on Craigie Laigh [NO 3124 7332] .

West of the Glen Doll Fault, Green Beds are found very close to the base of the formation in White Glen in the extreme south-west corner of the Ballater district. Traced north-north-east, these Green Bed lithologies transgress to higher structural levels and complex facies changes become apparent in the formation. The best exposed section occurs between [NO 2360 7265] and [NO 2341 7327] on South Craig in Mayar Glen. Similar but less comprehensive sections are found along strike on North Craig and in White Glen.

The Green Beds of the Ballater district are a varied group of rocks ranging from hornblende-bearing psammites to biotite-rich semipelites and both striped and massive amphibolites. This variation reflects their original diversity as generally turbiditic sediments with a variable volcanic component, and subsidiary basic lavas and possible subvolcanic sills.

The amphibolitic and metasedimentary rocks are interbedded on a millimetre to metre scale. This is particularly well seen on the rocky knoll 800 m west-south-west of Clova Hotel [NO 3202 7272], the lithological variation being picked out by distinctive ribbed weathering. The basic component consists of a massive, poorly foliated (S79333) to highly lineated (S79564), dark bluish green rock which is distinguished from non-green-bed type amphibolites by the presence of sporadic white quartz 'pebbles' 1 to 2 mm in diameter. In thin section, these quartz grains are highly strained, single or, more rarely, multiple grains which are contained by a finer grained intergrowth of quartz, plagioclase, hornblende and sphene. Additional minerals present include biotite and hematite (S79333) and epidote (S79564). As in the case of the Meall Dubh Metabasite Formation, epidote seems to be a diagnostic feature of metavolcanic rocks, and is not seen in the intrusive rocks.

The psammitic component is typically a pale greenish grey to buff-coloured gritty rock with disseminated pyrite cubes and, more rarely, clasts of hornblendic rock. It consists (S79565) of strained polycrystalline quartz clasts in a matrix consisting largely of granoblastic plagioclase. The rock is strongly altered; the plagioclase is intensely sericitised and the original mafic minerals are completely replaced by chlorite. The rock is rich in opaque minerals, both pyrite and magnetite, and also includes sphene and relatively coarse zircons. The assemblage on the knoll also includes bands of coarser and more schistose rock (for example (S79566)), comprising slightly greenish brown biotite intergrown with hornblende in rather discontinuous lithons, separated by granoblastic plagioclase and quartz.

The amphibole in the metasedimentary component of the Green Beds is normally green hornblende. However, the psammite (S79581) on Craigie Laigh [NO 3124 7332] additionally contains a colourless variety provisionally identified as cummingtonite. Also colourless anthophyllite (up to 2 mm long) is a major constituent of a gritty, garnet-bearing, quartzose psammite (S79571) notably deficient in micas which was collected from just below a metabasic sheet [NO 3170 7267]. Green Beds on The Strone (for example (S80937)) contain abundant poikiloblastic and skeletal grunerite and corroded mid-andesine plagioclase porphyroblasts in a foliated quartz-plagioclasebiotite-epidote matrix.

A large proportion of the Green Beds on South Craig [NO 23 73] comprise amphibolites, finely to moderately inter-banded with typically hornblendic semipelites and psammites. They are commonly migmatitic with abundant pegmatite segregation veins. The amphibolites may show lenticular banding and fracture patterns resembling those found in some basic lava flows. Psammites at [NO 2357 7297] are mid green-grey, gritty, non-migmatitic, volcaniclastic and micaceous, and show the typical brown-pink, carious weathering and massive, tough character of the Green Beds of the Southern Highland Group. These rocks (for example (S82921) and (S82922)) contain small lenticular quartz (and hornblende) clasts in a foliated quartz-plagioclase (andesine)-biotite-hornblende-garnet assemblage. The small euhedral garnets are intergrown with biotite. Minor clinozoisite is present and biotite shows partial alteration to chlorite. In the field, quartz veining and late stage cataclastic zones and mylonitic textures disrupt the earlier fabrics extensively. In adjacent gritty psammite units (for example (S82923)) the flattened lenticular quartz grains are present in a finer grained quartz-plagioclase (sericitised)-hornblende-ilmenite (sphene)-garnet assemblage. Minor epidote and chlorite are also present. The associated amphibolite units are commonly garnetiferous and typically (for example (S82919)) contain the assemblage green hornblende-plagioclase-quartz-biotite-garnet(magnetite). In parts of the area they are strongly magnetic with abundant scattered magnetite, but elsewhere sphene and/or ilmenite are present and the Green Beds have magnetic susceptibilities indistinguishable from background metasediment values.

Migmatitic striped amphibolites, biotite-rich semipelite and micaceous psammite, and hornblende-bearing psammite are interbedded with quartzose psammite and semipelite with minor calc-silicate lenses on the western side of Corrie Sharroch between [NO 2521 7474] and [NO 2553 7402]. Near to the southern limit of the section in Corrie Sharroch [NO 2544 7407] partly gritty, quartzose psammites contain epidote and quartz granules and thin layers. The sequence dips moderately towards the west-north-west but locally excellent grading shows that it is inverted and youngs towards the east-south-east. The epidote is presumably derived by erosion from underlying metavolcanic and volcaniclastic units.

Chapter 9 Metamorphosed intrusive igneous rocks

Nomenclature and chronology

The Dalradian of the Ballater district contains basic and granitic rocks (Figure 13) which were intruded at various stages during the Caledonian Orogeny and, as such, suffered differing degrees of deformation. The basic intrusions are divisible into older and younger suites on the evidence of intrusive contacts (Robertson, 1988), petrology, the degree of deformation and alteration, and comparisons with basic rocks within the Dalradian elsewhere in north-east Scotland. The distinction between the two suites is locally quite subjective; in several cases it proved difficult to categorise individual intrusions, mainly because in the field there is a convergence of appearance with increasing deformation of the younger suite. Rocks of the older suite are strongly deformed and are most certainly pre-tectonic to early tectonic in age. The younger suite is depicted as late (syn- D3) tectonic on 1:50 000 Series Sheet 65E Ballater, and was regarded as being coeval with Reid's (1919) Younger Basic suite. However, it is now recognised that the rocks, particularly the ultramafic ones, are petrographically similar to those of the Succoth–Brown Hill intrusion which Gunn et at (1996) have shown to be older than the nearby rocks of the (Younger Basic) Huntly mass. They are also petrologicaly closer to Reid's (1919) Older Basic than to the Younger Basic suite. Also some of the younger basic rocks in the Ballater district are affected by (D2–D3) shearing. Thus it seems (Table 1) more likely that these rocks are syntectonic rather than late tectonic as originally thought.

The granitic rocks were referred to collectively as Older Granites by Barrow and Cunningham Craig (1912) (compare with the post-tectonic Newer Granites), but on the basis of their relationship to the regional foliation comprise two distinct ages of intrusion. The older of these is the syntectonic Rough Craig Granite which was emplaced before the D2 deformation. The intrusion is undated, but the Rb–Sr isotope geochemistry of the Rough Craig Granite is similar to that of the Ben Vuirich Granite (Robertson, 1994) which has yielded a concordant U–Pb zircon age of 590 ± 2 Ma (Rogers et al., 1989) and which is now believed (Tanner and Leslie, 1994) to have been emplaced between the D1 and D2 deformations. The late-tectonic Cairn Trench, Hunt Hill, Cairn Inks and Driesh granites were intruded during the last stages of the D3 deformation. The Cairn Trench Granite has yielded a provisional age of 457 ± 5 Ma and initial 87/Sr86Sr of 0.7195 ± 2 (Robertson and Swainbank, 1990). The age is comparable to Rb–Sr ages for late to post-tectonic granites from the Buchan area such as Strichen and Longmanhill Granites (Pankhurst, 1974) and for a pegmatite in the Belhelvie area (van Breemen and Boyd, 1972).

Pre- to early-tectonic basic intrusions (Older Suite)

Metabasic bodies, formerly sills and discordant sheets and lenses of dolerite and basalt are the most abundant metamorphosed igneous rocks in the Ballater district but, because of the limited size of most bodies, are under represented on the 1:50 000 Series map. They occur widely throughout the lithostratigraphical succession, but are concentrated in the Glen Girnock, Queen's Hill, Water of Tanar, Pitlochry Schist and Rottal Schist formations. They are rare in the Blair Atholl and Islay subgroups, and in the Craig nam Ban Formation of the Easdale Subgroup, and are characteristically absent from the Longshank and the Cald Burn gneisses. Within the Glen Girnock and Queen's Hill formations there is a south-westerly decrease in the volume of metabasic material, whereas in the Rottal Schist Formation east of the Glen Doll Fault the volume increases to the south-west along strike. The only evidence that there might have been more than one intrusive phase is to be found on a glaciated slab above the Burn of Fialzioch [NO 2347 7715], where coarsely crystalline, green-black, foliated, highly garnetiferous amphibolite is in sharp contact with apparently later dark grey, fine-grained amphibolite.

The majority of the older suite intrusives, and certainly all of the larger units, occur as sheet-like bodies which, at outcrop scale, are concordant, but in a regional sense may be discordant to the Ethological layering in the surrounding metasedimentary rocks. The sheets are generally more resistant to erosion than the enveloping metasedimentary rocks and give rise to distinct ridges or spurs, for example above Strathgirnock [NO 306 731] and on Craig Maud [NO 237 767], or form impressive craglines such as on Cairn Inks [NO 30 72] to [NO 31 73]. The intrusions vary in thickness from a few centimetres to several hundreds of metres, but even within the thick units there may be numerous bands of metasedimentary material. Individual intrusions may be laterally continuous for several kilometres; the most persistent, which occurs within the Water of Tanar Limestone Formation, can be traced along strike for 22 km, and is commonly 500 m thick. The larger units may pinch out and reappear at different levels in the succession. This is particularly evident within the Water of Tanar Limestone Formation where either the metabasic unit described above or the psammite may lie adjacent to the metacarbonate rocks. In the Pollagach Gap area [NO 40 94] the metabasic unit may breach the limestone, and may occur below it; the structure of the area is at present poorly understood.

There are, however, some metabasic bodies which are markedly discordant for example at [NO 2338 8025]. Elsewhere, for example in the Water of Mark [NO 3818 8386] and near Mt Cholzie [NO 34948 831], minor metabasic bodies, a few metres thick at most, are seen to be slightly discordant in the hinge zones of major folds, but they are rotated into near parallelism with the metasedimentary rocks on fold limbs. These dyke-like metabasic bodies are rare, but widely distributed and volumetrically insignificant compared with the sheet-like bodies. Within the Glen Tanar Quartzite and Dounalt Limestone members, a number of the metabasic units form lenticular pods, such as around Little Hill [NO 37 85]. It is possible that some of these could have originated as crosscutting bodies. Other examples of possible dykes or flattened pipes occur in the Creag nam Ban Formation, for example on the northern slopes of Tom Bad a' Mhonaidh [NO 2883 9232] where the body is 30 m long and up to 3 m wide.

The metabasic rocks within the Glen Tanar Quartzite Member occur mostly as discrete pod-like bodies which, though quite large (for example around Little Hill [NO 37 85] ), lack the continuity exhibited by the sills within the Queen's Hill Gneiss Formation. However, a number of hornblende schist units to the east of Drumhilt, around [NO 638 805] are up to 20 m thick and can be traced along strike for up to 2 km. Some show millimetre-scale hornblendic and feldspathic striping which effectively defines the tectonic fabric.

The most extensive occurrence of the the older intrusive suite in the Ballater district is to be found in the lower part of the Rottal Schist Formation, south and west of Clova village. On the well-exposed craggy north-west face of Cairn Inks between [NO 306 731] and [NO 315 726] the hornblendic sheets comprise around 60 per cent of the outcrop which is 1.25 km wide. Individual intrusions range in thickness from 2 to 3 m to over 150 m, the larger ones typically having pinch and swell structures. In the same formation, west of the Glen Doll Fault, amphibolite metabasic bodies, mainly sills, lenses and some discordant sheets, are ubiquitous and notably abundant in Glen Doll, around The Strone, and east of Mayar in the valley of the Snow Burn. These bodies act as a locus for later folding, and typically show boudinage. Hence they both control and complicate the local and regional structural pattern, as well as disrupting the original stratigraphy.

The suite is dominated by hornblende schist, but also includes more massive, largely unfoliated amphibolite, metagabbro and actinolitic schist. Some bodies, such as the large lenticular sheet which is well exposed in the White Water between [NO 2444 7691] and [NO 2422 7704] range internally from coarse-grained unfoliated amphibolite to fine-grained highly foliated and lineated amphibolite. Typically the coarser grained and more mafic types show knobbly weathering. In the areas of highest metamorphic grade the hornblendic rocks are generally coarse grained and gneissose, and locally show migmatitic segregations. They vary from planar gneissose foliated to heavily agmatised (breccia-like) types, with perhaps 40 per cent leucosome in the rock. Agmatisation is, however, restricted to the large, coarse-grained metabasic sills, and seems to relate more to the original (igneous) grain size and composition rather than the metamorphic grade.

In the Queen's Hill Gneiss Formation, there are some units of a fine-grained, equigranular, garnetiferous hornblende schist which are slightly discordant to the litho-logical layering in the hinges of major folds. Typically they show marginal development of thick pegmatite lenses, for example at Knapps of Fee [NO 242 747]. The metabasic unit associated with the Water of Tanar Limestone Formation, though significantly larger and more extensive, is composed of a similar fine-grained, finely foliated, non-agmatitic rock studded with small garnets. However, the centre of the unit on Hare Cairn [NO 377 880], is coarser grained and unfoliated, and contains porphyroblasts of diopsidic clinopyroxene.

The older suite intrusives show a wide range of magnetic susceptibilities, with a significant number of highly magnetic specimens in the range 1 to 100 X 10−3 SI. As such they form several prominent magnetic marker horizons within the Queen's Hill Gneiss Formation. Even the least magnetic metabasic rocks tend to have higher susceptibilities than the surrounding metasedimentary rocks, thus forming distinct features on a MAGGRO map (Goodman et al., 1990).

Ultramafic rocks attributable to the older suite were recognised at two localities in the Queen's Hill Gneiss Formation, in and around Allt an t'Sneachda [NO 3317 8818] and [NO 3312 8773]. Both are at the south-eastern edge of a large amphibolite sheet, but are not considered to represent an early cumulate phase of the intrusion, since the regional stratigraphy indicates that they are at the top of the body.

With few exceptions the basic rocks are dominated by an amphibole-plagioclase (andesine) intergrowth with minor quartz. Within the suite, however there is considerable variation in the composition on both a macroscopic and microscopic scale; for example, the proportion of amphibole generally ranges between 50 and 70 per cent (but can be as low as 10 per cent). The suite also exhibits a wide range of textures, relating mainly to the intensity of fabric development. The fabric may be linear or planar, and is defined by an orientation fabric in the amphiboles and, to a lesser degree, by chains of sphene and opaque grains. In marked contrast to many of the syntectonic basic intrusions, igneous textures have all but disappeared through wholesale metamorphic recrystallisation. This is particularly evident in the plagioclase which forms mostly granoblastic polygonal aggregates commonly characterised by straight intergranular boundaries and 120° triple points. However, in a number of the more massive rocks the outline of certian of these aggregates suggests they could be recrystallised phenocrysts.

Generally the amphibole occurs as aggregates of stumpy to lath-like prisms or equant granoblasts of hornblende. Rarely, for example in thin section (S79323), larger relict grains which are distinctly poikiloblastic are present, and show an undulose extinction. In part, they have been replaced by a granular hornblende-plagioclase aggregate. Significant variations in hornblende colour which occur within the district appear to relate to metamorphic grade rather than to original composition. Thus, the metabasic rocks of the Glen Clova area within the kyanite zone are characterised by a blue-green variety, but moving up grade the colour changes through olive green to a distinctly brownish green ferroan pargasitic hornblende in sillimanite-grade rocks. It is not unusual to find blue-green hornblende mantling the olive variety, but its association with the brown variety appears to be restricted to later hornfelsed areas. Most of the metabasic sheets within Easdale Subgroup and older rocks contain blue-green and olive varieties, both of which have recrystallised to a distinctly poikiloblastic habit under the effects of thermal metamorphism.

Other amphiboles occur only rarely. For example fibrous pale green actinolite predominates particularly in Glen Clova (for example (S79314)), and cummingtonite was recorded in a specimen (S82034) collected from a colour-banded basic sheet, 500 m west-south-west of Camlet. The colour banding is due to alternating olive-green hornblende and cummingtonite-rich layers, although the two amphiboles may be epitaxially inter-grown. Garnets are common in the intrusive metabasic rocks within the Crinan and Tayvallich subgroups and the Southern Highland Group, but are absent from those within older stratigraphical units. They have developed preferentially in the more mafic types, and in rocks of sillimanite grade. Chemically they are almandines with up to 30 per cent grossular, a typical composition being Alm56Gr30Py7Sp7. They show evidence of only minor zoning and are apparently stable within the adjacent metamorphic assemblage. Biotite is present in some hornblende-poor metabasic rocks, commonly as a retrogressive phase. Sphene is an abundant minor component, forming irregularly shaped framboidal aggregates which in parts enclose or are rimmed by ilmenite. Opaque minerals, including pyrite, pyrrhotite, magnetite and ilmenite are common accessories; apatite, zircon and rutile may also be present.

Diopsidic clinopyroxene is present in metabasic rocks in a number of areas. For example, stable large, idioblastic porphyroblasts, up to several millimetres in diameter and with the composition Ca1Mg0.5Fe0.5Si2O6, are present within agmatitic metabasic rocks in the Cairn Leuchan–Drum Cholzie [NO 38 91] to [NO 38 47] area. Farther south, around the Scoube [NO 32 86], diopside is present as large, ragged porphyroblasts, speckled with inclusions. Elsewhere, for example near Fasheilach [NO 34 85], it is heavily retrogressed. Diopside was also recorded [NO 2935 9471], 550 m east of Abergeldie Cottages, in one of the few metabasic bodies recorded within the Creag nam Ban Formation. The rock (S82011) largely resembles a typical orthoamphibolite sheet, but includes irregularly shaped pale grey patches that are made up almost entirely of very coarse-grained (up to 5 mm) diopside along with minor amounts of plagioclase, clinozoisite and sphene. The diopside shows very localised alteration to hornblende, but in company with clinozoisite is extensively replaced by prehnite.

The presence of diopside in these rocks appears to be a product of original composition rather than metamorphic grade (compare with Baker and Droop, 1983). The mineral is restricted to the more mafic rocks which, with the exception of (S82011), have abundant garnet associated with the diopside. XRF analysis shows that these rocks are enriched in calcium relative to other early basic intrusives which would have aided the formation of diopside and the associated garnet. Ultramafic rocks are a minor component of the older suite. They are composed of olivine-pyroxenite, comparable with that in the syntectonic suite although more pervasively altered. They contain some remnant grains of ferroaugite, tremolite and olivine, but are chiefly composed of serpentine minerals, with abundant disseminated and dendritic pyrrhotite.

Geochemistry

Whole-rock X-ray fluorescence analyses have been made of five metabasic rocks from the Queen's Hill Gneiss Formation, including one discordant type, and two from the Water of Tanar Limestone Formation. Major and trace element values show that, like the Balnacraig metabasites, these rocks have tholeiitic affinities (Figure 11), and again in common with other Dalradian metabasic rocks (Graham, 1976b) lie close to the tholeiitic/alkaline join on the TiO2 v Zr/P2O5 plot. The trace element data for the discordant-type plots in the alkaline basalt field, although it still lies within the field of Dalradian metabasics defined by Graham (1976b).

Origin

It has been suggested (Harris and Pitcher, 1975) that the metabasic rocks in the Crinan Subgroup elsewhere in Scotland are high-level intrusions associated with the volcanic activity that took place during deposition of Southern Highland Group sediments. Certainly, there is a close spatial relationship between the older basic intrusives and the three principal volcanic horizons in the district that are essentially tholeiitic, two of which, the Meall Dubh and the Balnacraig metabasites, also have chemical similarities (Goodman and Winchester, 1993). The tholeiitic affinity is typical of the extensional regimes that prevailed in basin margin areas during deposition of the Argyll and Southern Highland groups.

In the Easdale, Crinan and Tayvallich subgroups there is a south-westerly decrease in the volume of metabasic rocks which, coupled with the paucity of recognisable dykes in the area, would seem to suggest that the volcanic centre lay in the Ballater area. Equally it appears that a second centre was established at the southern edge of the district during deposition of the Southern Highland Group.

Syntectonic basic intrusions (Younger Suite)

The rocks of the syntectonic basic suite have a more restricted distribution than those of the older intrusions (Figure 13), but where they do occur they tend to form somewhat larger bodies. The main concentrations occur on the west side of Glen Muick (Coyles of Muick) and on the north side of the Dee between Crathie and Ballater (Crathie, Corby Hall and Lower Glen Gairn bodies); much smaller bodies are present at a number of localities, notably at Balintober, on the Craig Megen summit ridge and in the upper reaches of the Water of Mark. They encompass a much greater range of rock types than the earlier suite, including metadiorite, metagabbro, amphibolite, plagioclase-amphibolite, hornblendite, pyroxenite and serpentinite.

In several areas, particularly around Crathie, these rocks are distinguished in the field from those of the older suite by a lack of foliation and relict igneous textures. However, syntectonic basic and ultramafic rocks from within the major shear zones normally possess a strong planar or linear fabric so cannot always readily be distinguished in the field. The problem is particularly acute in the area between Glen Gairn and the Tullich Burn, where there is strong shearing and both older and younger basic rocks occur in close proximity. They are, however, recognisable in thin section by mineralogy and partial preservation of igneous textures even in those rocks which have been substantially modified by shearing.

Coyles of Muick (sE and Uo)

The Coyles of Muick intrusion is depicted on 1:50 000 Series Sheet 65E Ballater as comprising both pre-tectonic to early tectonic ultramafic and late tectonic basic rocks, mainly because the ultramafic rocks display more intense alteration than the basic rocks, and the ultramafic rocks in the older suite tend to be concentrated in the Crinan Subgroup (Hawson and Hall, 1987). However, it now seems more likely that the rocks are coeval. Furthermore, they have many features (notably partial preservation of igneous textures in the basic rocks and a predominance of pyroxenites in the ultramafic rocks), in common with the Succoth–Brown Hills intrusion (Gunn et al., 1996) so the whole intrusion is considered to be part of the syntectonic suite.

The Coyles of Muick intrusion is situated to the southwest of Ballater on the the Glen Muick–Glen Girnock Burn watershed. Two small serpentinite bodies [NO 317 891] and [NO 3159 8964] on the Craig Megen summit ridge are also included in the following description, although they are divorced from the main body of the intrusion. The body is 5 km long and up to 1 km wide, and the northwest and south-east margins are defined by shear zones. At its north-eastern end, the body is truncated by the faulted edge of the Ballater Granite. Likewise its southwestern end is defined by an east-north-east-trending fault, although the body apparently tapers southwards. Much of the ground underlain by the intrusion is forested and exposure is poor, particularly in the north, so the external and internal boundaries were mainly established from ground magnetic surveys. Exposure improves considerably to the south of Cam a' Bhealaidh, particularly on the higher slopes and this is enhanced lower down the hill by more recent man-made exposure adjacent to forest tracks. Contacts between basic and ultramafic rocks are generally sheared. This is well seen in a track-side exposure [NO 3134 9272] where plagioclase amphibolite is separated from serpentinite by talcose shear zones. However, a contact between sheared metagabbro and hornblendite [NO 3252 9169] is quite oblique to the well-developed fabric in the basic rock (Plate 9). Also, at [NO 3159 8964] on the Craig Megen summit ridge, serpentinite appears to cut a 3 to 4 m-wide hornblende schist unit.

Because the better exposed ground tends to coincide with the outcrop of the north-west edge of the intrusion there is a preponderance of sheared rocks in exposures of the basic component, although the intrusion as a whole shows a range from highly sheared hornblende schist to massive unfoliated amphibolitised metagabbro. The latter is well exposed in a track side [NO 3331 9306] and consists of a very coarse-grained unfoliated aggregate of dark green hornblende and white plagioclase. The variable proportion of plagioclase makes the rock distinctly heterogeneous. It forms euhedral crystals up to 7 mm across, which occur either singly or in aggregates.

In thin sections of metagabbros (for example (S94407)), hornblende occurs largely as aggregates of granoblastic to prismatic crystals with variable grain size; in parts large subhedral plates mimicking the original clinopyroxene have survived and may ophitically enclose plagioclase. The plates show undulose extinction and partial recrystallisation where slightly deformed. Many of the larger plates contain finely divided opaque minerals in a gridiron pattern. These schillerised areas are noticeably idioblastic and do not always occupy the whole of the crystal being rimmed by inclusion-free hornblende. Coarser grained, lobate or rod-like magnetite is also included in places. An extreme development of this texture results in an intergrowth of comparatively fine-grained granular hornblende and coarse magnetite which, in places, is streaky or even dendritic, and may be accompanied by subhedral magnetites up to 4 mm long.

Plagioclase forms equant plates, up to 4 mm wide, which enclose small hornblende grains. Glide twinning, undulose extinction and partial recrystallisation with distinctly sutured intergranular boundaries are common in the larger crystals. Rarely, recrystallisation results in fine-grained annealed fabrics. Sericitisation of the plagioclase is widespread and locally intense; replacement by prehnite is rarer. Sphene occurs sporadically in poikiloblastic clusters which enclose fine-grained plagioclase and clinozoisite. More leucocratic plagioclase-dominated metagabbros were recorded [NO 3264 9171] and [NO 3252 9177].

Exposure of the ultramafic rocks is better than that of the basic rocks. They form some of the highest points on the ridge, for example Coyles of Muick (601 m) and Craig of Loinmuie (540 m), which, because of the high alkalinity of the soils, stand out as particularly verdant hills. The ultramafic bodies are typically lensoid with their longer axes usually parallel to the regional strike. They range in length from under 100 m to over 900 m, and in width from 30 m on Craig Megen to 600 m on the Coyles of Muick.

The bulk of the ultramafic rocks on the north-west side of Glen Muick consists of highly altered olivine-pyroxenite in which the original minerals have been replaced, to a large extent, by antigorite and tremolite. Antigorite replaces both olivine and orthopyroxene for example (S94441); where replacing the former it typically forms cellular aggregates that enclose abundant magnetite (characteristically in trails and etching outlines of individual cells) and, in places, relict olivine. The orthopyroxene, however, has altered to bastite showing schiller structure. Tremolite typically occurs as fibrous aggregates rarely enclosing small relict clinopyroxenes; it also forms large ragged plates. Clinochlore (chlorite) is the other principal alteration product, occurring mostly as bundles of crystals within the tremolite felt, and less commonly as narrow anastomosing veinlets. However, in thin section (S94444) tremolite is absent and clinopyroxene is replaced by large plates of clinochlore enclosing skeletal to poikilitic magnetite. Talc is present in a number of rocks, most notably in thin section (S94445) where it is the main pseudo-morph after orthopyroxene. Penninite and bowlingite also occur to a limited extent in some rocks. The Coyles of Muick serpentinite locally includes patches of carbonatised ultramafic rock (listwaenite). These comprise (for example thin section (S82153)) a granoblastic mosaic of dolomite/magnesite that encloses small grains of tremolite and slightly larger pools of partly serpentinised olivine.

A penetrative schistosity is common, largely defined by the alignment of tremolite and clinochlore. In thin section (S94441), trails of magnetite and a lensoid distribution of minerals enhance this fabric, which represents disrupted mineral layering/banding. More coherent mineral banding is evident in thin section (S82160), defined by alternating layers of fine-grained tremolite and coarse-grained tremolite with olivine and antigorite.

The ultramafic assemblage on the north-west side of Glen Muick also includes peridotites and olivine-free rocks. The former are dominated by olivine (for example (S82155)), but also include notably elongate tremolite prisms. Antigorite and clinochlore are only minor constituents and this rock is unusually fresh. The olivine-free rocks consist predominantly of tremolite with magnetite and clinochlore; thin section (S94443) additionally contains a few turbid clinopyroxene relicts, whereas (S82157) has irregularly shaped areas of talc that have probably replaced orthopyroxene.

The intensity of the shear fabric is very variable within the intrusion. It is particularly intense at the margins, but is also locally present within the intrusion. For example, thin section (S94440), which is mineralogically similar to, and was collected within 150 m of, (S94407) has a well-developed mineral lineation, and some of the plagioclase-rich domains are folded. The fabric is defined by a parallel alignment of hornblendes. More intense shearing of the metagabbro produces a noticeably streaky fabric that is generally more planar than linear. For example, thin section (S92972) is dominated by pale green to colourless, aligned and locally fibrous, actinolite prisms whose common alignment defines the fabric. The amphibole felt encloses foliation-parallel lenses and islands of heavily sericitised and in parts saussuritised plagioclase. The plagioclase appears to be metastable, most crystals have patchy extinction (reminiscent of unmixing), indistinct twin planes and sutured intergranular boundaries. The opaque minerals consist of subhedral to finely divided pyrite concentrated in areas where clinozoisite is developed. Deformation of the plagioclase-rich rock produces a texturally heterogeneous rock since the hornblendes have largely been replaced by granoblastic aggregates but the plagioclase is little deformed.

Crathie (Dc)

The Crathie amphibolite underlies an area of about 1 km2 on the north side of the Dee Valley, largely within grid squares [NO 26 95] and [NO 26 96]. It is well exposed on the southern flanks and summit plateau of Creag a' Chlamhain, and on Knock of Lawsie between [NO 259 963] and [NO 263 966]. Prior to the emplacement of the Abergeldie Complex, the outcrop of the body may have been more extensive since there are small roadside exposures on the A973 [NO 2728 9467], and xenoliths within granite below Balmoral Bridge [NO 2622 94942] and in dioritic rocks on the northern slopes of Tom a Chuir [NO 269 931].

The intrusion contains abundant xenoliths and rafts of metasedimentary rocks, mostly as strike-parallel lenses, in places traceable for several hundred metres. Their lateral continuity and orientation suggests that the intrusion comprises a series of sills with individual bodies ranging in thickness from 20 m upwards. They are broadly concurrent with the regional strike, but in the Crathie limestone quarry [NO 2689 9549] contacts are locally transgressive. The Crathie intrusion forms part of the roof zone of the Abergeldie Complex at [NO 2677 9544] and [NO 2646 9581] and it is punctured by stock-like bodies of quartz-diorite notably enriched in metasedimentary xenoliths. In many places the margins of the amphibolite form the focus for later intrusions; for example at the limestone–amphibolite contact on the eastern wall of the Crathie quarry there is a quartz-diorite vein up to 1 m wide, whereas nearby [NO 2660 9528] the vertical contact with quartz-diorite is occupied by a narrow feldspar porphyry dyke.

The Crathie intrusion consists of medium-grained, unfohated, feldspar-phyric amphibolite. The rocks generally have textures indicative of strong thermal metamorphism, being dominated by poikiloblastic to granoblastic inter-growths of hornblende and plagioclase. Except for some relict ophitic intergrowths and zoned plagioclase, primary igneous textures have been largely obliterated. However, in thin section (S82338) a number of squat to lath-like plagioclase (An45–52) blastophenocrysts up to 2.5 cm long survive, although extensive recrystallisation to granoblastic aggregates is apparent elsewhere in the thin section.

The amphibole is predominantly olive-green hornblende, characteristically occurring as a fine granoblastic aggregate, although in thin section (S82338) some crystals up to 3.5 mm have remained. These large crystals are much less poikilitic, have undulose extinction and incipient recrystallisation. Patches of a pale green to colourless amphibole may be present within hornblende. Clinopyroxene is present in thin section (S82346), largely as cores within hornblende aggregates, but in one part of the section it is concentrated along with hematite in a narrow vein. Biotite replacing hornblende was noted in a number of rocks, but quartz is rare. Anhedral to skeletal ilmenite is the principal accessory occurring within or surrounding sphene.

A most unusual variety of coarse-grained, iron-rich hornblendite forms part of the intrusion [NO 2651 9689]. The rock has a magnetic susceptibility of 180 x 10−3 SI, the highest recorded in the area. The most striking feature of the rock is the intensity of the colouration in the hornblende which ranges from deep yellow to turquoise. The habit of the mineral differs from the other amphibolites in that it is dominantly prismatic. Cores of clinopyroxene are also evident. Plagioclase is absent, and hornblende forms an ophitic-like inter-growth with andradite garnet. Garnet also forms vein-like aggregates which include hornblende and magnetite. Epidote and carbonate are common secondary minerals replacing garnet. Magnetite and sphene are the principal accessory minerals.

Corby Hall (Dg)

This small intrusion underlies the lower northern slopes of Creag nam Ban [NO 29 95] to [NO 30 95], east of Abergeldie Castle, and extends north beneath the flood plain of the River Dee. The southern boundary, where the basic rock abuts against metasedimentary rocks of the Creag nam Ban Formation, can be traced for over 500 m obliquely uphill (eastwards) from the track at the base of the slope; the boundary swings to the north-east around [NO 2998 9507] and appears to descend the hillside. The disposition of the bedding in the encircling metasedimentary rocks would seem to indicate that the basic rocks are in the core of an anticlinal structure, which may be the northerly extension of the F2 Camlet Anticline. Contact with rocks of the Abergeldie Complex on the north-west side is not exposed, but its proximity is inferred from the numerous granite veins within the basic rocks on the lower slopes.

The Corby Hall intrusion comprises a wide variety of rock types; it is dominated by medium- to fine-grained hornblende schist, in part feldspar-phyric, but also includes metagabbro and ultramafic rock. Included metasedimentary rock is restricted to a single body of psammite 150 m long and up to 30 m wide. Unlike the Crathie body this intrusion contains a pervasive foliation or schistosity which dips at 30 to 40° to the south-west. In the feldspar-phyric rocks the blastophenocrysts are aligned and flattened in the plane of the schistosity.

The dominant rock type (for example (S82021)) is a medium-grained amphibolite, showing a rather ragged, poikiloblastic intergrowth of pale greenish brown hornblende and plagioclase with a distinctive foliation defined by common alignment of the amphiboles. An amphibolite (S94406) collected from the south bank of the Dee, 80 m west of Corby Hall, contains irregularly shaped pale green patches composed of medium- to coarse-grained clinopyroxene and heavily sericitised plagioclase, and devoid of hornblende. Close by, there is a small exposure of coarse-grained, unfoliated, hornblende metagabbro (S94405) which retains much of its primary ophitic texture. Although no contact with amphibolite is visible, the metagabbro is assumed to be part of the same body.

Ultramafic rocks are restricted in occurrence, and are found only [NO 2991 9540] south of Corby Hall. The exposure has two differing lithologies; thin section (S82018) shows a schistose felt of pale green to colourless actinolite, scattered through which are ragged grains of olivine, magnetite and hercynite. Thin section (S82019) consists predominantly of much larger crystals of tremolite (in part replaced by aggregates of phlogopite), which in turn enclose granular olivine, hercynite and magnetite. In comparison with (S82018), hercynite forms larger grains, but olivine is smaller, less common and generally quite turbid.

Lower Glen Gairn (Dc)

There is a notable concentration of younger suite basic sheets intruding the upper part of the Blair Atholl Subgroup, and the Islay and Easdale subgroups, in a 6 km-wide north–south zone which straddles the lower part of Glen Gairn. Many of these sheets are near-concordant although discordant contacts between the sheets and host metasedimentary rocks, locally with screens of host rock arranged en échelon, confirm the intrusive origin of these rocks. Cross-cutting relationships between different metabasic lithotypes together with local preservation of possible chilled margins between them (Tocher, 1961) indicate repeated intrusion of metabasic sheets. The abundance of sheets generally increases towards the east, such that a large body of metabasic rocks underlies 4 to 5 km2 between Hill of Candacraig [NJ 34 00] and Creagan Riabhach [NO 37 99]. This body is probably about 1 km thick, but as intrusive contacts are rare and metasedimentary rocks absent from the area, it may represent a single intrusive sheet or coalescence of sheets. Screens of schistose fine-grained amphibolite within medium-grained largely unfoliated metabasic rocks on Craig of Prony [NO 353 988] may represent relict chilled margins or fragments of older basic rocks. To the north of this, on the Hill of Candacraig there is very clear evidence that these younger basic rocks cross-cut rocks of the older suite (Robertson, 1988) (Plate 10).

These basic sheets encompass a large range of litho-types, from even-grained to feldspar-phyric variants, and from unfoliated to highly schistose rocks. The rocks are generally fine grained where they are highly deformed. Some are metagabbros which have preserved igneous features, but the majority are pervasively recrystallised to amphibolite. At several localities in the vicinity of Creagan Riabhach [NO 3745 9891]; [NO 3712 9897], the basic sheets exhibit a pronounced layering due to millimetre- to centimetre-scale alternations of hornblendic and feldspathic domains which may be a primary igneous feature.

Feldspar-phyric varieties, thought to be derived from porphyritic dolerite protoliths (Tocher, 1961; Boutcher, 1963), occur mostly in sheets less than 10 m thick. They contain feldspar aggregates up to 10 mm across which in places comprise up to 50 per cent of the rock. The distribution of the feldspar aggregates varies over short distances with patches of feldspar-phyric amphibolite within homogeneous amphibolite and vice versa. Gradations from feldspar-phyric to even-grained rock occur both parallel and perpendicular to sheet margins. The least strained rocks contain tabular plagioclase megacrysts, but with increasing deformation these have recrystallised to plagioclase aggregates, some of which are pseudomorphed by white mica (Boutcher, 1963). They occur within a hornblende-plagioclase groundmass which is indistinguishable from the overall appearance of the even-grained amphibolites. Sphene, opaque minerals and quartz occur as accessories.

On the Hill of Candacraig, amphibolite sheets up to 40 m thick preserve lenses of coarse-grained metagabbro wrapped by schistose amphibolite within anastomosing shear zones. One kilometre north-west of Creagan Riabhach [NO 3623 9995] coarse-grained metagabbro is cut by a 5 cm-thick pegmatitic metagabbro with grain size up to 25 mm. The rocks are deformed by shear zones, some less than 1 cm across. In thin section some of the metagabbroic rocks contain augite in the form of cores to hornblende crystals and as discrete grains within coarse-grained plagioclase.

Amphibolites in an area extending west-north-west from south of Peter's Hill [NO 372 998] to the eastern slopes of Hill of Candacraig [NO 349 005] are cut by veins and patches of grey to pink leucotonalite ranging in thickness from a few millimetres to more than 25 cm (Boutcher, 1963). They vary from pink leucocratic veins with sharp straight contacts, to diffuse areas and 'ghost veins' which are only slightly more leucocratic than the host amphibolite. The veins are composed largely of quartz and turbid plagioclase, with variable but subordinate amounts of hornblende, biotite, sphene and opaque minerals. The veins do not occur outwith the amphibolites and appear to predate the deformation of the host. Gradational contacts suggest that the quartzofeldspathic material was not intruded after consolidation of the metabasic rocks, which led Boutcher (1963) to suggest that the leucotonalite veins represent a late differentiate of the basic magma.

Balintober

Coarse-grained, fresh ultramafic rock (U), and a hornblende-plagioclase rock (H), which can be coarse and equigranular or finer grained and foliated, are exposed near Balintober Farm [NO 36 93]. The rocks occur as pods, up to 150 X 50 m in size, strung out along the sheared junction of the Balnacraig Metabasite Member and the Queen's Hill Gneiss Formation and parallel to the strong foliation in the surrounding Dalradian rocks. The absence of a thermal aureole in the country rocks suggests that, like the nearby Coyles of Muick intrusion, these rocks have been emplaced by shearing and are now separated from their original envelope. Specimens of the Balintober suite have relatively high magnetic susceptibilities, in the range 0.1 to 10 X 10−3 SI.

The ultramafic component is a pyroxenite, containing abundant enstatite with some tremolite and fayalitic olivine, the latter being 2 to 3 mm in diameter. Typically, it forms whaleback outcrops for example [NO 364 937] which show the knobbly weathering surface and weathered yellow crust of ultrabasic rocks, but the rocks themselves are usually very fresh, with little or no serpentinisation. This contrasts with the ultramafic rocks of the older suite which are all extensively serpentinised. At one exposure [NO 3649 9378] primary igneous decimetre-scale compositional layering is seen in ultramafic rocks, and a similar structure may be present in some of the hornblende-plagioclase rocks. This layering dips to the north, and is parallel to neither the margins of the ultrabasic pod, nor to the regional Dalradian foliation.

The hornblende-plagioclase rock for example at [NO 367 938] contains coarse-grained hornblende and andesine feldspar, with biotite, bright green chlorite, some quartz and disseminated opaque minerals. Despite its massive appearance this rock still has a faint compositional layering in places. It also grades from a coarse equigranular rock with a distinctly igneous appearance to a sheared, foliated type, the fabric being parallel to the shear fabric of the surrounding Dalradian metasedimentary rocks.

Glen Mark

The Glen Mark metadiorite (sH) is a concealed body, being overlain by up to 2 m of peat, till and deeply weathered rock. However, diorite veining and a thermal overprint are evident in adjacent country rocks, and the body gives rise to a prominent magnetic anomaly running along the east side of Black Hill of Mark [NO 32 81]. The anomaly, which can be traced for 6 km, cuts across the regional strike and is associated with north–south faulting. Trenching established that the concealed rock is a sheared diorite, mainly formed of hornblende (60 per cent) and plagioclase (30 per cent), with some quartz, orthoclase, biotite and opaque minerals, and with an anastomosing shear fabric. Recorded magnetic susceptibilities of 1 to 100 SI confirmed that the metadiorite is the source of the magnetic anomaly.

The metadiorite is divided into a number of slices by north-north-east-trending faults, some of which are marked by breccia carrying epidote, garnet and baryte. The eastern margin of the metadiorite is probably fault bounded, and the position of this boundary as determined from the magnetic model agrees with the observed boundary in nearby pits to within ± 5 m. Regionally, the country rock foliation appears to swing into a north–south orientation approaching the meta-diorite, suggesting large scale ductile movement, prior to the brittle faulting. It is likely that the metadiorite was originally intruded as a dioritic sheet into an active, broadly D3, shear zone, which acted as a conduit for intrusion. Further shearing led to amphibolitisation of the diorite (Kneller and Leslie, 1984) and the formation of shear fabrics in the metadiorite and the surrounding Dalradian. Later still, under a more brittle regime, there was brittle faulting along the same north–south line, which thus has a long-lived history of movement. The north–south structure can be traced as a brittle fault to the north where, in places, it forms the eastern boundary of quartz-diorites marginal to the Lochnagar Complex.

At a number of localities to the north-east of Black Hill of Mark for example at [NO 33 83] north of Sheiling of Mark, [NO 36 90] on the west flank of Am Mullach and [NO 38 90] to the south of Head of Garbh Choire there are smaller but essentially similar magnetic anomalies which are also associated with north or north-north-west faulting. There is no exposure along any of these anomalies, but on the basis of their magnetic characteristics, they are considered to be due to the presence of small bodies of sheared metadiorite, presumably of the same structural age and geological history as the Glen Mark metadiorite.

Syntectonic and late-tectonic granites (Older Granites)

Barrow (1893) identified a total of six intrusions of muscovite-biotite gneiss in the Glen Clova–Glen Esk area. Harry (1958) subsequently recognised that two of the areas of muscovite-biotite gneiss comprised migmatitic gneisses of metasedimentary origin, but confirmed that the others were mostly intrusive microcline granites. Robertson (1991) has shown that in the ground between Glen Clova and Glen Esk the granites are of two ages, syntectonic (pre-D2) represented by Rough Craig, and late-tectonic (D3) represented by the Cairn Trench and Hunt Hill granites.

The major occurrences of granitic rocks on the south-west side of Glen Clova are recognised as being part of the late-tectonic suite. In addition to the centres on Cairn Inks and south of Driesh recognised by Barrow, granite and pegmatite veins south-west of Cadham, in Corrie of Farchal and in Glen Doll are also part of the late suite.

Syntectonic granite

Rough Craig

The Rough Craig Granite (GRC) is exposed on Rough Craig [NO 353 727] and Brown Holm [NO 343 736], with smaller bodies to the east of Rough Craig around [NO 357 732] and [NO 369724] Similar granite sheets are recorded from the crags enclosing Loch Brandy, above the Water of Unich [NO 367 781] and from as far north as the Burn of Doune in Glen Mark [NO 393 848].

It is a grey foliated granite composed of quartz, plagioclase, microcline, biotite and muscovite with apatite and zircon. The rock is typically medium grained, but in the Burn of Doune around [NO 393 848] the typical granite is accompanied by gneissose muscovite-biotite granite with feldspar augen up to 1 cm in size. Gneissose variants, which are well exposed on the south-east spur of Rough Craig [NO 353 727] and on Brown Holm [NO 344 735], are marked by the occurrence of regularly spaced (3 to 5 mm) biotite lamellae separated by quartzofeldspathic layers which produce a prominent banding. In places the banding is irregular, such that lenses and patches of granitic leucosome occur within biotite-muscovite-rich melanosome (Plate 11). It is not clear whether such areas represent partially assimilated xenoliths or areas of lower strain where the regular banding has not developed.

Elsewhere, the gneissose granite passes laterally into foliated granite. In the absence of contact or structural relationships, weakly foliated variants are indistinguishable from the Cairn Trench Granite. The gneissose banding in the granite is locally enhanced by subsequent migmatitic segregation which produced concordant leucosomes up to 10 mm thick. The leucosome is distinguished from the pre-existing banding by the presence of pink feldspar segregations, an increase in grain size of the quartzofeldspathic minerals, an enhancement of the mafic selvages marginal to the leucosomes and the lensoid form of the segregations. Subsequent migmatisation of the granite is shown by pegmatitic segregations along the axial surfaces of D3 fold structures.

The contacts of the main body of the Rough Craig Granite are not exposed. However, exposures of granite are seen within 1 m of host metasedimentary rocks on the south-east spur of Rough Craig [NO 3543 7264]. Here both the contact and the tectonic fabric of the granite are concordant with the regional S1/S2 fabric in the metasedimentary rocks, which together with the outcrop pattern implies that the main granite body forms a near concordant sheet.

Textures of the granites reflect their complex tectonic and metamorphic recrystallisation history. Although the rocks mostly appear medium grained, many thin sections exhibit a wide range of grain size. Recrystallisation to fine-grained aggregates is abundant at the margins of larger quartz and feldspar grains. Quartz may occur in the form of ribbons indicative of strong deformation, whereas other rocks only display a mica preferred orientation and no segregation. Grain size reduction has occurred adjacent to some mica laminae.

Several rocks contain microcline and zoned plagioclase megacrysts typically 4 to 6 mm in size, which poikilitically enclose other feldspars, biotite, quartz and in rare cases, muscovite. These may represent relicts of original magmatic textures. The rarity of included muscovites suggests either that it crystallised as a late magmatic element or there was little primary muscovite. Scattered large quartz grains or aggregates of grains, up to 4 mm in size, are typically flattened in the plane of the fabric. Mica laminae wrap around the large feldspars. Muscovite occurs both as laths and large plates and augen up to 3 mm in size; the augen have rounded corners. Many muscovite augen are spatially associated with areas of grain size reduction adjacent to mica-rich laminae, suggesting that their development is at least, in part, related to shear deformation. However, some muscovite, discordant to the mica fabric, occurs with the quartzofeldspathic lithons. Biotite exhibits a wide range of grain size up to about 1 mm; some is replaced by muscovite. A single subhedral garnet, 1 mm across occurs in specimen (S81858) from the south-east spur of Rough Craig.

Late-tectonic granites

Cairn Trench

The Cairn Trench Granite (GCT) crops out in the southeast of the Ballater district, principally in the area bounded by Loch Wharral [NO 36 74], the Water of Saughs [NO 40 75] and Kennel Burn [NO 39 71]. The granite occurs as both generally discordant veins of granite and pegmatite, and as sheets and larger bodies of granite. The latter are well exposed in the area to the south of Cairn Trench and in the vicinity of Kennel Burn.

The granite is typically pink to grey, medium to coarse grained, and composed of quartz, plagioclase, micro-dine, muscovite and biotite. In places, it is sparsely porphyritic with microcline megacrysts up to 10 mm in size. Biotite megacrysts have also been recorded, apparently spatially related to the presence of biotite schlieren. Rafts of metasedimentary rock and mafic schlieren are locally abundant. In the vicinity of Cairn Trench, exposures show gradation from homogeneous granite, through granite with mafic schlieren, to biotite-rich rocks resembling migmatitic palaeosomes, possibly derived from semipelite protoliths. Psammite xenoliths show little evidence of assimilation. In places the granite becomes pegmatitic. However, at one locally [NO 3789 7397], discrete pegmatitic sheets, 5 to 20 cm thick and dipping about 40° north, occur within the granite at intervals varying from 30 cm to several metres.

The granite is locally foliated with the fabric defined by the preferred orientation of micas. Both foliated and unfoliated granites crop out in the Kennel Burn area, although relationships between them are not seen. A loose block of weakly foliated granite, found to the southwest of Loch Wharral [NO 3537 7369], contains xenoliths of more melanocratic foliated granite, although it is not known whether such foliated granites are of Rough Craig or Cairn Trench affinity.

The Cairn Trench Granite is modally indistinguishable from the Rough Craig Granite. Texturally, there are also many similarities, including the occurrence of perthitic orthoclase and microcline, zoned plagioclase and muscovite megacrysts, together with a large range in grain size and abundant myrmekite. However, although the Cairn Trench Granite displays extensive recrystallisation, it lacks most of the tectonic features (such as gneissosity) common to the Rough Craig Granite.

Cairn Inks

The Cairn Inks Granite vein complex is located roughly 2 km west-south-west of Clova village within Grid Squares [NO 29 71]; [NO 29 72]; [NO 30 71] and [NO 30 72]. It comprises a sheeted granite and pegmatite complex intruded into the metasedimentary and metabasic rocks forming the lower part of the Rottal Schist Formation. The granite and pegmatite sheets outcrop over an area of about 2 km2 at the southern edge of the Ballater district; about 4 km2 crop out to the south in the Kirriemuir district (Sheet 56E). Because of poor exposure the nature and form of the individual granite and pegmatite units is not precisely known. The most extensive exposure, centred on [NO 3022 7181] is 40 m wide and can be traced in a north-east direction (parallel to the strike of the regional foliation) for only 200 m. The contact with the country rocks is not seen.

Satellite to the main complex, there are a number of sheets and veins on the north-east face of Cairn Inks and on the eastern shoulder of Hill of Strone. The most extensive of these between [NO 3055 7265] and [NO 3066 7302] can be followed for 500 m through a vertical height of over 300 m. The sheets range in thickness from less than a metre to 30 m. On a regional scale they generally parallel the main schistosity/foliation (S1/S2). However, locally they may be seen to cut across the fabric, for example the essentially flat-lying sheet on the lip of the corrie between Cairn Inks and Hill of Strone is oblique to the principal fabric and also transgresses the base of the Rottal Schist Formation. The intrusions in the area south-west of Cadham [NO 331 725] are also considered to be part of the complex even though they lie 1.5 to 2 km east of the main body.

Discordant contacts are more commonly encountered in pegmatite sheets. For example [NO 966 7316] in Corlowie, a 10 m-wide unit can be followed up the western wall of the corrie for over 100 m through essentially flat-lying semipelites. The cross-cutting nature of the pegmatites is apparent even on the 1:50 000 scale, as at Craig Duff [NO 326 720], where a 10 m-wide sheet can be followed obliquely across a large metabasic body and into the country rocks.

Petrographically the complex consists mainly of pinkish white to buff or, more rarely grey, medium- to coarse-grained muscovite-biotite granite and pegmatitic muscovite granite. Much of the granite has an obvious foliation defined by the alignment of mica. In more extreme cases (for example thin section (S80257)) the rock has developed a gneissose banding made up of alternating quartzofeldspathic and micaceous layers. The granite (for example (S81052)) consists essentially of quartz, K-feldspar, oligoclase, muscovite and biotite, with garnet in places, and with accessory apatite and zircon. In most of the rocks examined, patchily perthitic microcline is the main feldspar phase, but orthoclase is also present and a few (for example (S81050)) are dominated by plagioclase. Muscovite in general is commoner and coarser grained than biotite; the latter forms ragged flakes with a colour range from red-brown to deep brown which may show partial replacement by chlorite. Garnet appears to be a primary phase of the rocks, although in thin section (S81049) a crystal within microcline has a reaction rim of sericite.

The distribution of pegmatite within the sheets is quite erratic, although there seems to be a concentration on the eastern edge of the complex between [NO 308 723] and [NO 311 716] . Essential mineralogy of the pegmatites is quartz, plagioclase, microcline and muscovite with sporadic garnet. These rocks are mineralogically very heterogeneous; for example, one of the sheets on Craig Duff which consists largely of quartz, feldspar and muscovite is represented solely by white quartz at its western end.

Driesh

The Driesh granite vein complex is centred on Hunt Hill [NO 262 720] some 4 km west of Cairn Inks. Like the Cairn Inks granite complex it occurs on the southern edge of the Ballater district. However, it is considerably smaller with an outcrop area of no more than 1.5 km2 of which roughly 20 per cent lies within the district. As at Cairn Inks establishing the full extent of the complex is hampered by a limited exposure. The principal occurrences are on the summit plateau and steep eastern slopes of Hunt Hill. From there the complex can be traced eastwards through float, and small exposures in the Burn of Eskielawn [NO 2650 7219] and [NO 2656 7215], to the Shank of Driesh where there are sporadic exposures on the track between [NO 2734 7239] and [NO 2725 7175]. Outlying occurrences were also recorded on the southern slopes of Driesh [NO 2692 7308] and Little Driesh [NO 2645 7266].

The Driesh Granite strongly resembles that at Cairn Inks, comprising a sheeted complex within metabasic and minor metasedimentary rocks of the Rottal Schist Formation. The sheets are up to 30 m thick and consist largely of medium- to coarse- grained, foliated muscovitebiotite granite which is locally pegmatitic. Where exposed at [NO 2640 7200] on Hunt Hill the granite is white and more aplitic, and is characterised by wispy micaceous parting. This exposure is also notable in that the granite is demonstrably cut by pegmatitic muscovite granite.

The sheets consist mostly of quartz, microcline, possible orthoclase, plagioclase, muscovite and biotite. However, a specimen (S81043) collected from a 20 cm-thick concordant vein on Little Driesh consists almost entirely of an interlocking intergrowth of quartz and andesine with no visible K-feldspar. The rock is characterised by euhedral garnets up to 5 mm across, is noticeably deficient in micas, particularly biotite, but is enriched in clinozoisite.

Hunt Hill

The Hunt Hill Granite (Gu) crops out in the Water of Lee–Water of Unich area of Upper Glen Esk Grid squares [NO 36 78]; [NO 37 77]; [NO 37 78]; [NO 37 79]; [NO 37 80]; [NO 38 79] and [NO 38 80] and in the Burn of Doune area in Glen Mark (Grid square 39 84). It occurs mainly in the form of low-angle, subconcordant sheets apparently less than 20 m in thickness, although steeply inclined discordant sheets are also recognised. They mostly occur intruding the Glen Tanar Quartzite Member, but also occur in the Longshank Gneiss Formation. Pegmatitic leucogranite sheets on the north slopes of Glen Lee around [NO 386 818] are up to 6 m thick and occur as en échelon lenses, wrapped by the regional S1–S2 foliation. Small-scale pinch and swell structures occur at granite contacts suggesting that the lensoid form of the sheets has resulted from boudinage.

The granite is composed of a white to locally pink, commonly pegmatitic leucogranite containing quartz, plagioclase, microcline, muscovite (locally up to 15 mm across), tourmaline, small amounts of biotite and, in places, garnet. Like the Cairn Trench Granite, it contains abundant myrmekite and a range of grain sizes resulting from partial recrystallisation. In general, the Hunt Hill Granite is texturally heterogeneous, ranging from micro-granite through granite to pegmatitic granite over short distances. Metasedimentary xenoliths and micaceous schlieren are widespread and a banding defined by biotite-rich laminae is developed in places.

Minor occurrences

Elsewhere in the south-western corner of the Ballater district, older granite occurs in lenticular veins generally subparallel to the regional foliation. For example, exposures are found as subcrop and felsenmeer (block fields) on the east shoulder of Mayar [NO 2431 7387] and [NO 2423 7387], where a foliated, pale fawn pink, fine-grained muscovite granite occurs at the amphibolitesemipelite boundary. In Mayar Glen [NO 2357 7295], adjacent to Green Beds, two near-vertical, white to pale pink, foliated, fine- to medium-grained muscovite-biotite granite sheets, 3 and 4 m wide, lie slightly discordant to the metasedimentary layering. The sheets have pegmatitic margins but show little leucosome or pegmatite development. The granite (S82920) consists of K-feldspar, plagioclase (oligoclase-andesine), muscovite, biotite and chlorite, with accessory apatite, ilmenite and rare zircon. Microcline and orthoclase crystals, up to 3 mm long, occur in about equal proportions and perthite is rare. The average grain size is about 1 mm across, but marginal to the larger crystals are zones of recrystallised quartz and feldspar, locally with myrmekite development. Zoning was not observed in the plagioclase (compare with the post-tectonic granites), and the straggly small biotites are commonly replaced by chlorite or overgrown by muscovite. Muscovite plates, up to 1.5 mm long, define the pervasive foliation and adjacent flakes typically lie in optical continuity suggesting that it recrystallised during deformation and metamorphism.

A distinctive white, massive, fine- to medium-grained muscovite-leucogranite sheet, 3 m thick, intrudes micaceous psammites close to the top of the Queen's Hill Gneiss Formation in upper Corrie Fee. At exposure [NO 2439 7494]; [NO 2444 7505] and [NO 2432 7469] it attains a thickness of 10 m. It consists ((S82926), (S82949)) of poikilitic perthitic K-feldspar (1 to 2.5 mm across and commonly showing microcline twinning), low andesine and quartz, in a finer grained matrix of quartz, plagioclase, muscovite, altered biotite and antiperthite. Accessory ilmenite and apatite are present. The phenocrysts are strained, bent and fractured, with strain perthite abundant in the K-feldspars. Quartz is recrystallised, both to subgrains and to fine-grained aggregates, but the large strained grains show deformation lamellae in parts. Muscovite both replaces biotite and forms late-stage plates and aggregates. Sericitisation of plagioclase is seen.

Geochemistry of the syntectonic and late-tectonic granites

The Rough Craig and Cairn Trench granites are geochemically very similar with a small range in both major and trace element compositions; minor differences are restricted to slightly lower V, MgO and TiO2 and slightly higher Ce and Th in the Cairn Trench Granite for a given Zr concentration ((Figure 14)a–e). Both granites are corundum- and hypersthene-normative. CIPW normative compositions plot close to the three-phase cotectic line at 5 kb water pressure in the granite ternary system (Figure 15).

Little geochemical data for other 'older' granites in the East Grampian area have been released in the literature. O'Brien (1985) presented a limited number of analyses for the Aberdeen and Strichen granites; both are considered to be Ordovician in age (Kneller and Aftalion, 1987; Pidgeon and Aftalion, 1978) and therefore possibly coeval with the Cairn Trench Granite. These data reveal the much greater range in composition of the Aberdeen and Strichen granites compared with the Clova granite. Strichen is anomalous in its Y content which is higher than the other three granites, whereas both Aberdeen and Strichen have higher Th and Ce ((Figure 14)d-e). Strichen has high Ba/Sr and low Rb/Sr compared with the other three granites.

Chapter 10 Structure

The metasedimentary and to a lesser extent the meta-volcanic rocks of the Dalradian succession have a compositional banding (S0) which is taken to reflect original lithological variation. This is generally regarded as bedding in the metasedimentary rocks except where there is obvious transposition or metamorphic segregation. A few examples of sedimentary structures have survived, for example graded and cross-bedding in the Cairnwell Quartzite Formation, graded bedding in the Rottal Schist Formation and slump structures in the Crathie Schist and Limestone Formation. However, the biggest aid to unravelling the structural history of the district is the detailed knowledge of the lithostratigraphy.

Deformational structures

The most obvious and widespread evidence of deformation is the presence of a penetrative schistosity or gneissosity, except where subsequent migmatisation or metamorphic recrystallisation has been particularly intense. In pelitic and semipelitic rocks the fabric is largely defined by parallel arrangement of mica; in some psammitic lithologies the fabric is a spaced cleavage; within metabasic rocks it stems from a common alignment of amphibole which may result in a linear rather than planar fabric. The fabrics are all related to folding which is the other main visible evidence of deformation.

There is ample evidence that, like the Dalradian elsewhere in Scotland (Harris et al., 1976; Harte et al., 1984; Mendum and Fettes, 1985), these rocks have been subjected to at least four episodes of deformation. Evidence of multiple deformation is seen at a number of localities where two or even three fabrics are present. For example, in White Glen [NO 2364 7200] blocky quartzose psammite with a prominent early spaced cleavage, which dips more gently than bedding, is overprinted by a second more steeply dipping cleavage and a third, weak, discordant, spaced cleavage. Outwith these areas it is not always possible to distinguish between fabrics, particularly S1 and S2. In some areas, it is evident that the fabric is composite. Likewise the fold chronology is based on a limited number of exposures showing refolded folds (Plate 4); in most of these only two phases are present, but north of Sgor an h-Iolaire [NO 3017 9411] an F2 closure is clearly refolded by F3 and F4 folds. This evidence, coupled with distinctive orientation and characteristic style (for example whether isoclinal, tight, close or open) of the various fold episodes, and correlations with structures in adjacent areas is extrapolated to folds elsewhere in the Ballater district.

D1 deformation

This is the least well defined of the four deformations, largely because its effects have been extensively overprinted by subsequent events. The principal evidence of the early deformation is a planar fabric which is generally bedding-parallel. In parts of Glen Doll this simply reflects the original compositional banding, but elsewhere, for example the south-west side of Glen Clova [NO 3298 7184], the early structure is more obviously tectonic, being a spaced cleavage which parallels the lithological layering and is extensively obscured by the dominant S2 spaced cleavage. Similarly, a pervasive cleavage on the Creag nam Ban ridge can be followed round the nose of the major F2 syncline. A tectonic fabric defined by micaceous laminae, in places isoclinally folded intrafolial to the S2 spaced cleavage, is recognised within quartzose lithons south of Delnabo [NO 3044 0053] and on the Milton Spur [NO 3128 9900].

The early fabric is not universally concordant with the lithological layering. For instance, south-east of Sgor an h-Iolaire [NO 3070 9222] a well developed fabric which predates F2 folding is oblique to bedding. Similarly in the Green Beds of the Lochanluie area, thin parallel quartz layers, which are thought to represent an S1 pressure solution cleavage, are discordant to both S2 schistosity and bedding.

The S1 fabric is believed to result from the formation of the Tay Nappe, a regional-scale recumbent anticline which closes to the south-east. Most, if not all of the Dalradian rocks in the Ballater district lie, like those in the Braemar district (Upton, 1984), on the lower limb and on what was the flat section of this structure (the 'Flat Belt' of Johnstone, 1966), but because of subsequent movements, particularly folds of the third deformation, over much of the district, So, S1 and S2 have moderate to steep dips. Probably the only unmodified relict of the 'Flat Belt' is to be found on the south-west side of Glen Clova immediately west of the Farchal Fault.

Except for a small area on the eastern edge of the district, the Dalradian succession consistently youngs to the south-east. However, if the dip of the rocks is taken into consideration then it is apparent that in the northwest the succession is right-way-up, whereas south-east of the Coyles of Muick Shear Zone the Dalradian rocks are for the most part inverted (Figure 16). It could be argued that the juxtaposition of the two limbs is due to movement along the Coyles of Muick Shear Zone but it will be shown later that the difference in attitude is more likely to have stemmed from F3 folding.

A break in the general south-east-younging pattern occurs in the area of Glen Mark, where the Tarfside Psammite Formation reappears below the Southern Highland Group and forms part of a 'right way up' succession extending east through the Tarfside area of Glen Esk; Harte (1966, 1979) believed this to form the upper limb of the Tarfside Nappe, a structure considered to underlie the Tay Nappe. A number of features led Harte to postulate that the two nappes were juxtaposed by a major tectonic slide which he called the Glen Mark Slide; these included the termination of the Green Beds to the north-east of Glen Clova, together with 'reversals in bedding' in that area recognised by Anderson (1942), the occurrence of high-strain rocks in Glen Mark and lithological changes in the Southern Highland Group farther to the south-east. The slide was inferred to occur in glens Lee and Mark, extending north from Loch Lee [NO 410 790] to the south-western corner of the Kincardine Granite at Knowe of Crippley [NO 403 808].

Facies changes within the Argyll and Southern Highland groups, including the abrupt termination of the Green Beds, occur across a north-west-trending lineament (Chapter 4) and relate to deposition rather than later tectonics. Rocks on Craig Damff [NO 380 792], on the east-facing crags of Hunt Hill [NO 383 804], on Cairn of Camlet [NO 398 815], in parts of Glen Lee to the south of Wolf Craig and in the Burn of Doune around [NO 401 837] are marked by planar and linear high-strain textures including mylonitic fabrics. These are spatially related to the boundary between the Tarfside Psammite Formation and the Longshank Gneiss Formation rather than the postulated slide and, as such, may be related to the contrasting ductility of these units. In the absence of direct evidence for a tectonic slide, it is suggested that the inverted and right way up successions are juxtaposed by a major tight to isoclinal fold. However, minor folds do not constrain the age of folding. The absence of a change in vergence of F3 fold structures across the contact suggests that the fold is a F1 or F2 structure. Since major F2 folds elsewhere in the Ballater district do not cause widescale repetition of the stratigraphy, it is more likely that this is an F1 structure.

F1 minor folds are present in places (Plate 12), for example they are clearly seen in the hinge zones of later F2 folds on the east side of Glen Muick. These are usually rootless isoclinal folds, no more than 5 to 10 cm in wavelength, and they are intrafolial to later fabrics. In F2 hinge zones, it is possible to see that the Si fabric associated with the isoclines strikes counterclockwise of the S2 fabric. Thus on Creag Mullach [NO 412 952] and Pannanich Hill [NO 392 939] the strike of the Si fabric is north-south, whereas that of the S2 fabric is north-north-east.

The intrafolial folds deform a fabric which is not bedding, but an early tectonic fabric. In metasedimentary and most metabasitic rocks this is a schistosity, though it may contain stringers of quartzofeldspathic material. In some metabasites within the F2 hinge region south of Craig Vallich [NO 3602 9090], an evenly spaced fabric is deformed into early tight folds marked by quartzofeldspathic material. This is reminiscent of a tectonic striping or spaced cleavage being folded by the F1 folds. This earlier fabric probably relates to an early phase in the protracted Tay Nappe deformation event.

In places, the early fabric is formed of leucocratic stringers, but these may be mimetic after a more schistose earlier fabric. The segregations are never large enough to be termed 'migmatitic', though leucosome is commonly developed along the limbs of the isoclines, on the later fabrics. The main period of migmatisation therefore probably postdated the F1 folding.

Tadesse (1991) identified a zone of high strain within the rusty quartzite and psammite unit at the top of the Creag nam Ban Psammite Formation which he termed the Creag nam Ban 'Slide'. The zone is characterised by a strongly penetrative mylonitic fabric and a complex fold pattern not seen in adjacent rocks, which includes refolded folds with attenuated limbs and thickened hinge zones. The close association of the slide with the rusty quartzite psammite unit could well explain the lateral impersistence of this unit. The slide is folded by a major D2 fold pair and cut by the S2 fabric and hence is regarded as being a product of the D1 deformation.

D2 deformation

The main effect of the D2 deformation was a transposition of S0 and S1 into the plane of S2 such that the dominant fabric in many areas is largely an S1/S2 composite. Over much of the Ballater district S2 dips moderately to steeply to the north-west, although local variations resulting from F3 folds occur. For example, north-west of the Coyles of Muick Shear Zone the dip of S2 is predominantly to the south-east. In most semipelitic and pelitic rocks the second fabric consists of a penetrative schistosity. In the coarser grained and higher grade rocks it takes the form of a gneissosity, which in stromatic migmatites is paralleled by the leucosomes.

There is no evidence that the D2 deformation causes significant repetition of the stratigraphy. Despite the widespread presence of D2 fabrics and minor folds, major F2 folds were only identified in the ground east of Glen Muick and in the Abergeldie area (Figure 16). In the former area these structures comprise two synform–antiform pairs, with north-east-trending axial plane traces that can be followed for 7 to 9 km. These F2 folds are asymmetric; the geometry is best illustrated by the outcrop pattern of Cairn Leuchan, Craig Dearg and Drum Cholzie metabasic rocks, but is also seen in outcrop-scale folds for example at [NO 3606 9091]. The folds have long north-west-dipping limbs and short horizontal or south-south-east-dipping limbs, and verge towards the north-west. The folds plunge to the south-west, except where they are affected by later folding. The main regional foliation is axial planar to the F2 folds and dips to the north-west. The long limbs of the F2 folds are transposed into this foliation, but the short limbs of the folds retain a more primitive fabric. There is a strong mineral lineation down dip on S2 which is associated with the F2 folding.

Two major F2 folds, the Creag nam Ban Syncline and the Camlet Anticline, are recognised in the Abergeldie area. The axial plane trace of the syncline has a north to north-west trend and can be followed through closures in the So and S1 fabrics for up to 3 km northwards from Camlet [NO 308 931] along the Creag nam Ban ridge. At the northern end of the ridge it is truncated by the Abergeldie Igneous Complex, but a probable extension exists in the north-north-east-trending synformal closure on the south-west flank of Geallaig Hill [NO 29 98].

The Creag nam Ban Syncline is an upright, tight to isoclinal, asymmetrical fold. Congruent minor folds have steeply to moderately dipping axial planes and the mean fold axis orientation dips 54° towards 210° according to Tadesse (1991). The Camlet Anticline is a much more poorly defined structure, largely because of a F4 overprint. Immediately west of Camlet the axial plane trace is north–south and parallel to that of the Creag nam Ban Syncline, but traced to the north the two appear to diverge and the trace of the anticline swings towards the north-north-east. The anticline is a much more open structure than the syncline.

The Creag nam Ban–Camlet fold pair have several aspects which are atypical of F2 folds elsewhere in the district and which probably reflect the effects of later deformation. In particular the north–south alignment of the axial planes contrasts with the more usually encountered north-east trend, and the openness of the anticline is a feature not encountered elsewhere. Also, the southwest plunge of the parasitic minor folds contrasts markedly with the east-south-east regional trend for F2 folds north-west of the Coyles of Muick.

In places, for example on Creag Mullach [NO 412 952] and near Allt Cholzie [NO 349 881] there is interference between D1 and D2 structures. The hook-like structure in metabasic rocks in the hinge of the F2 Cairn Leuchan fold [NO 38 92] appears at first to be an interference pattern. Close inspection shows no change in the geometry of the S1/ S2 intersection in this area, and it seems more likely that irregular geometry of the metabasic body causes the unusual outcrop pattern. Similarly the great thickness of distinctive aluminous pelites between the Creag Dearg and Drum Cholzie metabasic rocks [NO 35 88] is due to thickening in the F2 fold hinge, rather than the presence of the aluminous pelites in an F1 closure.

Coyles of Muick and Tom-Candacraig Shear Zones

Two zones of intense shearing, the Coyles of Muick Shear Zone and the Tom-Candacraig Shear Zone, have been recognised in the Ballater district (Figure 16). They form part of a major zone of discontinuities running from Glenshee to the Moray Firth coast. Fettes et al. (1984) suggested that the southern part of the lineament included the line of the Glen Doll Fault running along the eastern side of the Lochnagar Granite. However, Goodman (1994) concluded that there was an overall southerly decrease in the intensity of shearing such that the Coyles of Muick Shear Zone represented the southern extremity of the lineament. No evidence of shearing was recorded in the area of the Glen Doll Fault during this survey.

The Coyles of Muick Shear Zone can be followed for nearly 7 km north-eastwards from the Craig Megen ridge [NO 31 89] to where it is truncated by the Ballater Granite. It is 2 to 3 km wide, although the boundaries particularly on the south-east side can be quite diffuse. The zone, which is near-vertical and parallel to the regional strike, is recognised in upper Easdale and lower Crinan subgroup rocks and, to a lesser extent, in outcrops of syntectonic basic bodies. Within the Coyles of Muick Shear Zone individual shear horizons vary in width from narrow spaced slip surfaces, lined by fine-grained phyllosilicate minerals and ferromagnesian solution seams, to metre-wide mylonitic and phyllonitic schists (Tadesse, 1991). The frequency of slip surfaces and widths of the mylonitic and phyllonitic horizons increases towards the centre of the zone and there is a concomitant recrystallisation of porphyroblasts and decrease in grain size. S- and Z-shaped and north-westerly verging asymmetrical minor folds are associated with the shear zone deformation, and the mylonitic fabric carries an oblique northeast-plunging stretching lineation. In the Creag Liath Pelite Member and the Meall Dubh Metabasite Formation there are pervasive post-tectonic recrystallisation, grain coarsening and annealing microtextures. In contrast, the microtextures developed in rocks of the Queen's Hill Gneiss Formation are similar to those found in sheared granites with the annealed textures never attaining the protolith grain size.

The Coyles of Muick Shear Zone had a sinistral sense of movement (Tadesse, 1991). This is evident in the broad asymmetric configuration of the main regional fabric, together with the north-easterly thinning and disappearance of the Meall Dubh Metabasite Formation and to some extent the Creag Liath Pelite Member. Additional evidence is provided by a range of commonly encountered asymmetrical structures including plagioclase augen, pressure shadows around garnet and staurolite, composite planar fabric relationships and extensional structures.

The Tom-Candacraig Shear Zone extends for 1.5 km from the east-north-east-trending fault between Hill of Candacraig and Craig of Prony to the northern edge of the district. At its southern end, the zone has a (foliation-parallel) north-north-west alignment, but as it passes out of the district it swings round to the north, thus obliquely cutting across the lithostratigraphical boundaries before coalescing with the South Cabrach Shear Zone 2.5 km to the north at [NJ 340 030]. The Tom-Candacraig Shear Zone is 50 to 70 m wide (Tadesse, 1991) and affects units of the Creag nam Ban Psammite and Glen Girnock Calcareous formations together with pre-tectonic to early-tectonic and late-tectonic basic intrusions. The zone was described by Tadesse (1991) as comprising mylonitic bands, up to 1 m wide, broadly parallel to the regional structural trend, with north-north-west-trending strike, and a steep dip towards the east-north-east. Locally the foliation is sinuous and encloses bodies of undeformed basic material. The mylonitic bands may be affected by overturned minor folds which plunge subhorizontally to the north-east and have a south-easterly sense of vergence. The Tom-Candacraig Shear Zone in contrast to the Coyles of Muick Shear Zone has a dextral sense of movement.

On the evidence that the shear fabric partly obliterated the S2 fabric and locally affected the post-D2 late-tectonic basic intrusions, Tadesse (1991) considered that it was part of the D3 event, although his plate 5.16c appears to show mylonitic rocks folded by regional F3 folds, and Robertson (1988) has shown that units forming part of the syntectonic basic suite were emplaced in metasedimentary rocks which were already strongly sheared. On balance, it seems most likely that much of the shear/mylonitic fabric developed late in D2, after the peak metamorphic mineral assemblage had formed, and that further, but more limited, movement occurred during or after emplacement of the syntectonic basic rocks, giving rise to a local S3 fabric which preceded the main regional D3 event.

D3 deformation

Early D3 structures

The relationship of the localised D3 shearing with the earlier main event can only be established north of the Dee. This is because S2 and the local S3 are largely coplanar and can be readily separated only where rocks of the syntectonic basic suite are present. In practice, the distinction is on the basis that some of the metabasic sheets, which are concordant with S2 but cross-cut F2 folds, have a later, margin-parallel, penetrative schistosity.

The S3 schistosity is not ubiquitously developed within the syntectonic basic intrusions. Many carry a schistosity only within discrete shear zones that range from a few centimetres to several metres in thickness. Such shear zones are particularly well displayed within coarse-grained metagabbroic rocks. The sinuous nature of the mylonitic fabric within some of the syntectonic basic bodies in the Coyles of Muick area is a reflection of the heterogeneity in the local strain orientation, and does not indicate that it formed other than synchronously with the more planar fabric on the margins.

Linear fabrics are more widely developed, so that many metabasic sheets carry only a hornblende lineation or a stretching lineation defined by 'rodded' plagioclase aggregates. L3 is coaxial with F2 in the host metasedimentary rocks. The stretching linear fabric is most intensely developed within the syntectonic basic bodies and especially south of the Cam Dearg Fold trace, indicating that these rocks were particularly susceptible to stretching, possibly because of the absence of a pre-existing planar anisotropy.

Carn Dearg Fold

Over much of the district the strike of the lithological layering approximates to the north-east or 'Caledonian' trend which prevails throughout much of the Scottish Dalradian, but in the ground north of Ballater the attitude abruptly changes to north-north-west. As the Caledonian trend is only regained in the Lecht area, over 20 km to the north-north-west, this represents a major change in the orientation of the Dalradian rocks. The change in orientation occurs across the Carn Dearg Fold, the west-south-west trace of which extends from the Sluggan Burn around [NO 375 998] to Craig of Prony, and along the Cam Dearg watershed towards Geallaig Hill. Lithological layering is generally near vertical to the north of this structure, but to the south dips moderately south-east. Since there is a comparable change in the orientation of S2 and the local S3 across the axial plane trace, Tadesse (1991) considered that it was a D4 structure. D2 and D3 linear fabrics plunge moderately towards the south-east throughout the area so there is no change in orientation across the Cam Dearg Fold trace. Furthermore a stereoplot of regional fabric orientations from both north and south of the fold trace form a girdle about an axis which plunges gently to moderately southeast (Tadesse, 1991; Boutcher, 1963) and is therefore broadly coaxial with F2 and L3. The coaxial form of F2 and L3 means that the Carn Dearg fold could be attributed to either D2 or D3 phases of deformation. However, since the younger basic sheets are apparently folded by the structure the balance of evidence suggests it is a (local) D3 upright antiformal structure.

Main D3 structures

Folds attributed to the third phase of deformation are less widespread than those of D2, but this phase gives rise to the extensive steepening and overturning of the Dalradian rocks (Figure 16). For example the right way up sequence in the north-west part of the district forms the upper limb of a major south-east-closing, recumbent D3 structure, the Craig Megen Fold, whose north-east-trending axial trace runs along the watershed between Glen Girnock and Glen Muick. This structure was not previously recognised, possibly because the steepening of the lithological layering and associated fabrics were thought to be due to movement along the Coyles of Muick Shear Zone, and there are few if any F3 minor folds in the axial trace zone.

The main visual indication of the existence of the Craig Megen Fold is the progressive change in dip of the layering and the S2 and (local) S3 fabrics, from steep to the south-east on the north-west side of the ridge, through vertical to moderate to steep to the north-west on the south-east side. Other evidence is provided by the change in plunge of the F2 fold axes from south-west on the lower limb to south-east on the upper, although the vergence remains consistently to the north-west. Parasitic F3 minor folds are comparatively rare in the northern part of the district; they show a general south-easterly sense of vergence with the shorter limb being steeper. On the south-east side of the fold trace they verge in the opposite direction. Evidence from a prominent F3 minor fold on the Creag nam Ban ridge [NO 3023 9460] suggests that the axial plane of the Craig Megen Fold may dip westwards at up to 25°.

The third phase of folding is seen mostly in rocks occurring in the southern half of the district, and is generally associated with a zone of gentle north-west dip in the foliation. The F3 folds are open to tight, near-recumbent structures which range in scale from crenulations in micaceous lithologies to large-scale fold structures controlling the disposition of the lithostratigraphy. The axial traces of many of the major F3 folds can be traced for several kilometres despite the poor level of exposure. For example, although the bulk of the succession in the southern part of the district is inverted, the Southern Highland Group rocks on the south-west side of Glen Clova between the eastern shoulder of Bassies and the Farchal Fault retain a right-way-up succession, being on the western limb of the F3 Corlowie Fold. Also, the F3 fold pair between Hunt Hill [NO 380 805] and Drumhilt [NO 353 805] can be followed for nearly 10 km to the north-north-west to where the folds are truncated by the eastern edge of the Lochnagar Granite complex. The F3 structures refold the earlier planar fabrics, isoclinal fold structures and extension lineations. For example, the F2 fold axes have a southerly plunge in the north-east part of the district, but are reorientated to a more variable northerly plunge in the Capel Burn area [NO 30 78], to the south-west of a major F3 hinge zone.

F3 folds in the southern half of the district are markedly asymmetric and show a sense of overthrusting to the north or north-west (Plate 13). Regionally, long fold limbs dip gently to moderately north-west and short limbs, which may be up to 4 km long, dip gently to moderately to the south-west or south. On a more local scale there are significant variations from this regional pattern.

For example, on Wolf Craig [NO 38 82], where F3 fold closures are defined by infolds of Longshank Gneiss Formation within the Tarfside Psammite Formation, all fold limbs dip gently to moderately north.

D3 structures dominate the deformational history of the rocks in south-west of the district, in the area bounded by the Glendoll Fault and the Lochnagar Granite. The D3 event there was particularly complex with recrystallisation and new metamorphic mineral growth occurring during deformation, some local thrusting and strong attenuation on the limbs of F3 folds accompanied by a locally penetrative fabric. F3 folds are ubiquitous in this area, notably where the rock types are striped alternations of psammitic and semipelitic lithologies. Kilometre-scale examples are shown by the outcrop pattern of thick amphibolitic sheets, notably in the Craig Maud area of south-west Glen Doll. The F3 folds typically have gently south-west- to south-plunging axes, and range from open to very tight depending on the type of lithology, overall structural position and original geometry of the banding. Their axial planes dip moderately to steeply to the west-north-west. Related lineations plunge gently to the north-north-east. In Glen Doll [NO 2377 7725] the minor folds verge to the east-north-east and have strongly attenuated limbs where S3 and S1/ S2 become subparallel. Such attenuated limbs climb up the fold profile like a shear/thrust zone.

North-east of Glen Clova, most F3 folds plunge gently to the west, although north-west- or north-north-west-plunging structures occur together with west-plunging structures in places. The relationship between the north-north-west- and west-plunging structures is not clear; they may represent a conjugate set of folds.

The axial plane orientations of the F3 folds show some systematic regional variations. In the area immediately west of the Glen Doll Fault, F3 axial planes trend fairly consistently north-north-east and dip moderately to steeply west-north-west, whereas farther west, in Glen Doll and near Loch Esk they dip to the west-south-west, and even to the south-west at steep angles. The Glen Doll Diorite and movements along or across the Glen Doll Fault have undoubtedly affected F3 orientations locally. Around Glen Clova the axial surfaces have quite gentle dips; on the south-west side these are mostly to the northeast or, more rarely, to the west, whereas to the northeast the orientation is to the north-west or north. In the north-west of the district, axial planes of the F3 folds dip gently to the west. Axial planar fabrics to F3 folds are generally restricted to crenulation cleavages in more micaceous lithologies; more penetrative S3 fabrics are rare but were recorded at a number of sites within the Corby Hall basic intrusion [NO 29 95].

The vergence of small-scale F3 folds confirms that the prominent knee-bends in the outcrop pattern of Crinan and Tayvallich subgroups and Southern Highland Group rocks in the south-east of the district are controlled by large-scale F3 fold structures. Such large-scale F3 folds have not been recognised elsewhere in the Southern Highlands. Their occurrence in the Glen Clova–Glen Mark area is thought to reflect ductility contrasts at the boundary between Argyll Group and Southern Highland Group rocks. In effect, during deformation the Tarfside Psammite Formation acted as a resistant block which was 'wrapped' by the ductile more pelitic lithologies of the Longshank Gneiss and Rottal Schist formations.

D4 deformation

There is little evidence of any major penetrative post- D3 folding in the Ballater district, although local areas of low dip in earlier structures, such as around Balnacraig [NO 35 92] and in Glen Muick appear to be the result of a late deformational event. However, small-scale, late, upright folds occur throughout the district and are notably concentrated in certain areas such as between Corrie Fee [NO 245 749] and Dounalt [NO 243 763] adjacent to the Glen Doll Fault. There, the F4 structures are mono-formal, generally stepping down to the east or south-east, in places tightening to close folds. Typically, they show near horizontal or gently west-dipping limbs that have undulating minor folds and steep or near-vertical limbs. Orientations of fold axes show a considerable spread, but with two notable concentrations plunging at very shallow angles towards about 010° and 215°. Axial planes dip moderately to steeply west-north-west, but again show a spread in orientation. Several axial planes dip west-southwest at varying angles (shallow to moderately steep). The fold axis and axial plane orientations are influenced by the pre-existing fold patterns, in particular the F3 fold pattern. This is reflected by the spread in F4 axes in the south-west.

In the south-east, small-scale monoclinal folds or kink bands occur on the Craigs of Cormaud in Glen Lee [NO 381 815] and in Glen Mark. They have subhorizontal plunges with axial surfaces striking north-north-east. They consistently step down any particular layer to the east-southeast. A probable F4, upright, open fold, with a north-trending axial trace in the Burn of Longshank [NO 354 779] is recognised on the basis of the orientation of S1–S2. Also in the Burn of Longshank [NO 3601 7821], a crenulation cleavage dipping steeply north may be related to D4 although this is not clear cut. On the south-west side of Glen Clova, the F4 fold axes plunge subhorizontally to the north-east or east. The axial planes are invariably steep to subvertical, the dip being to the south-west or south-east.

Upright minor folds are common on the southern part of the Craig Megen ridge (Plate 7) where they are seen to fold the shear fabric in the Queen's Hill Gneiss Formation. In this area the F4 axes plunge moderately to steeply to the north-east, the mean being 64° to 060° (Tadesse, 1991). Axial planes are predominantly vertical with a north-east-trend. On the ridge immediately north of Camlet, somewhat larger scale F4 folds warp the S1–S2 fabric on the south-east limb of the Carnlet Anticline. The axial trace of these folds is north–south, with F4 axes plunging moderately to the south. In places, the F4 folds have an associated fracture cleavage or crenulation cleavage, the latter containing a retrogressive mineral assemblage that includes muscovite in the metasedimentary rocks and biotite in the metabasic rocks.

Chapter 11 Regional metamorphism

In the late 19th century, George Barrow first drew attention to the zonal arrangement of metamorphic minerals while carrying out the Geological Survey's primary survey of the south-eastern Highlands. He interpreted the zones as indicating a progressive northerly increase in the degree of metamorphism related to the emplacement of the Older Granites (Barrow, 1893). It is now recognised that the metamorphism is not related to the older granites. Some of the zone boundaries are lithologically controlled and hence are not strictly speaking isograds, nevertheless the concept of the Barrovian zones as indicators of prograde regional metamorphism remains unchallenged and is a testimony to Barrow's excellent pioneering work. The results of Barrow's initial work in the Glen Clova–Glen Esk area, published in 1893, established the existence of staurolite, kyanite and sillimanite zones. Barrow (1912) subsequently added garnet, biotite and digested elastic mica zones, and extended the coverage to include all the ground between the Highland Boundary Fault and the Dee valley as far west as the Linn of Dee. The area described in this memoir thus provided much of the material used by Barrow to erect the concept of metamorphic index minerals.

Rocks occurring in the ground between the Dee and the Banffshire coast were subsequently shown by Read (1952) to possess a somewhat different assemblage of metamorphic minerals characterised by the presence of andalusite and cordierite, with kyanite absent. These index minerals, which formed at lower pressures than the Barrovian assemblages, are the product of what has become known as Buchan metamorphism. An effective boundary between the two systems is the andalusite–kyanite isograd.

The Dalradian rocks in the Ballater district are predominantly within the sillimanite zone as mapped by Barrow although, because of the southerly decrease in metamorphic grade, rocks in the extreme south-east lie within the kyanite and staurolite zones (Figure 17). Barrow (1912) also identified a wedge of kyanite-bearing rocks which extended along Glen Girnock for 5 km from the eastern edge of the Lochnagar Granite. More recently, on recognising regional andalusite + garnet in the Camlet area, Chinner and Heseltine (1979) concluded that the andalusite–kyanite isograd runs approximately east-northeast from where the Girnock Burn cuts the eastern edge of the Lochnagar Granite. Thus while the bulk of the Dalradian rocks in the Ballater district have mineral assemblages typical of Barrovian metamorphism, those in the north-west part are of Buchan character.

Barrovian metamorphism

Textures and metamorphic reactions in the southern part of the Ballater district have been described and discussed by Chinner (1960, 1961, 1965, 1980), Harte and Johnson (1969), Harte and Hudson (1979), Dempster (1985) and McLellan (1982). For this generalised account of Barrovian metamorphism in the Ballater district the rocks are considered on a zone by zone basis. The description refers to semipelitic and pelitic rocks unless stated otherwise since these lithologies contain a mineralogy most responsive to the conditions of metamorphism.

Staurolite Zone

Staurolite zone rocks are restricted to the Rottal Schist Formation and only occur in the extreme south and south-east of the district. They are typically medium grained and comprise the assemblage:

quartz + plagioclase + biotite + muscovite + garnet + staurolite + tourmaline + ilmenite + magnetite

A penetrative S2 schistosity is produced largely by the preferred orientation of biotite and muscovite. Muscovite-rich laminae and layers enhance the fabric in some specimens. Garnet, staurolite and locally biotite and plagioclase porphyroblasts, where present, act as loci for F3 microfold hinges, indicating that they grew before the third deformation (D3). Porphyroblasts, notably biotite which forms augen structures, are wrapped by the muscovite fabric. However, clearly defined quartz pressure shadows adjacent to porphyroblasts are not recognised.

Staurolite and garnet are almost invariably poikiloblastic, and contain inclusions of quartz, plagioclase and minor biotite. These are generally finer grained than equivalent minerals outwith the porphyroblasts, which Harte and Johnson (1969) interpreted as evidence of matrix grain coarsening after porphyroblast growth. In places, euhedral garnets are included within staurolites. Quartz and feldspar inclusions typically define an internal fabric which may be regional Si, since it is the earliest fabric preserved and definitely predates the regional S2 fabric. On the evidence of inclusion assemblages and the character of internal fabrics, particularly their deformation and relationship to the external schistosity, both staurolite and garnet show evidence of post-D1 but pre-D2, and syn- D2 growth.

Where present, biotite porphyroblasts contain prominent pleochroic haloes and commonly occur adjacent to, or include, tabular and skeletal ilmenite. Commonly, such ilmenites are speckled with fine-grained quartz inclusions. Neither the haloes nor the ilmenite are features of biotite laths in the rock matrix.

Plagioclase porphyroblasts in the form of anhedral augen, up to 2 mm in size, occur in some specimens. Many contain quartz inclusions; garnet and biotite inclusions are less abundant and staurolite and tourmaline are recognised in only one specimen (S81864).

Tourmaline grains, pleochroic from colourless to bluish green, are included within staurolite and occur within the swathes of muscovite.

Biotite porphyroblasts with associated ilmenite commonly occur adjacent to staurolite, although in many cases with intervening quartz ((Plate 14)a). Relicts of garnet occur in a few biotite porphyroblasts. These features may indicate the reaction:

chlorite + muscovite + garnet = staurolite + biotite + quartz (see Winkler, 1979) (Reaction 1)

Alternatively, the quartz rim may reflect instability of the staurolite–biotite assemblage. Several staurolites contain ilmenite inclusions, similar to those within the biotite porphyroblasts, particularly those staurolites which overgrow biotite with included ilmenite. This suggests that locally staurolite has replaced biotite. In places, ilmenite is included within both staurolite and garnet which is itself included within the staurolite, as in thin section (S81864). This may indicate that both staurolite and garnet have developed at the expense of the biotite porphyroblasts. The textural similarity between garnet and staurolite porphyroblasts together with the inclusion of euhedral garnets within staurolite, suggest coeval growth of the two minerals. Staurolite relicts also occur within a few biotites. The origin of these is not known. The contrast between the relicts of garnet and staurolite in biotite, combined with the occurrence of euhedral garnets included within staurolite, that appears to overgrow biotite, indicates the existence of more than one phase of garnet and staurolite growth. Many garnetbiotite grain boundaries are embayed and biotiteilmenite porphyroblasts occur as protrusions in garnet. The cores of some garnets are replaced by biotites, with or without muscovite and quartz.

Plagioclase and garnet show embayed contacts with muscovite laminae, and plagioclase is commonly altered, particularly to sericite, along its cleavages. This is consistent with the approximate inverse relationship between the abundances of muscovite and plagioclase. The origin of the muscovite is not clear, although it has at least in part overprinted biotite as indicated by their intergrowth. However, there is no evidence for a significant release of mafic components from such a replacement. It may indicate hydration of the metamorphic assemblage. Since the preferred orientation of muscovite largely defines S2, its formation would be late D2. Staurolite commonly preserves straight grain boundaries against muscovite.

Muscovite 'herring-bone' textures and granoblastic polygonal textures in quartz lithons in the hinge zones of D3 crenulations ((Plate 14)b) show that some muscovite and quartz recrystallised after D3. The occurrence of garnet and staurolite relicts within biotite porphyroblasts hints at an earlier phase of porphyroblast growth.

Kyanite Zone

Rocks containing kyanite without fibrolitic sillimanite are restricted to a narrow zone, less than 1 km wide (Figure 17), straddling the boundary between the Longshank Gneiss and the Rottal Schist formations.

Mineral assemblages are broadly similar to those within the staurolite zone with the addition of kyanite. Observations from a small number of specimens suggest that there is an inverse relationship in the abundances of kyanite and staurolite (see Harte and Hudson, 1979). Staurolite is absent from some rocks close to the margin of the sillimanite zone. Although poikiloblastic and inclusion-free staurolites may occur in the same rock the relationship between the two types is not known.

Inclusions within poikiloblastic garnets tend to be coarser grained than those seen in the staurolite zone although they are still commonly finer grained than the matrix. Inclusions define both straight and sinuous internal fabrics. Some straight inclusion trails are parallel to the external fabric, whereas some sigmoidal and crenulated trails are continuous, at garnet margins, with external fabrics. Textural relationships are, therefore, similar to those in the staurolite zone, indicating growth during D2. Some garnets are partially replaced by biotite, muscovite and, in places, plagioclase. Garnet cores were more susceptible than rims to replacement resulting in the development of 'atoll' textures.

Biotite porphyroblasts only occur close to the boundary with the staurolite zone. Elsewhere, they are absent, although matrix biotite is prominent where associated with kyanite.

Kyanite typically occurs in the form of anhedral to subhedral laths and plates, locally up to 6 mm long. Some kyanites contain inclusions of quartz and opaque minerals as well as biotite. These are particularly abundant in a magnetite- and tourmaline-rich rock (S81885), in which the inclusions define an internal fabric, discordant with the external schistosity. Kyanite is also found within muscovite and quartz, and occurs on a macroscopic scale within quartz segregations. Several examples of clusters of microscopic kyanites within quartz crystals have been recorded. Kyanite and biotite also occur included within plagioclase, although the textural relations between these minerals are not clear. Elsewhere, there are examples of kyanite including muscovite, and kyanite included within muscovite that contains biotite lamellae with diffuse margins. The latter texture is similar to that seen in the staurolite zone; it may indicate hydration, with replacement of biotite by muscovite after kyanite growth. A few kyanites contain overgrowths of acicular kyanite, suggesting that there were two distinct periods of kyanite growth.

Close to the outer margin of the kyanite zone, kyanite first appears within or close to staurolite ((Plate 14)c). The presence of biotite, quartz and muscovite suggests that the following reaction was important in the generation of kyanite:

staurolite + muscovite + quartz = kyanite + biotite + H2O (McLellan, 1985) (Reaction 2)

This reaction occurs at lower temperatures in Mg-rich rocks than in Fe-rich rocks and as such the appearance of kyanite does not define an isograd. Chinner (1965) therefore defined the kyanite isograd as the plane along which kyanite coexists with biotite of a specific composition, namely MgO x 100/ (MgO + FeO) = 51.7.

A few kyanites occur in association with biotite at garnet grain boundaries although there is no evidence for new garnet growth in this area. Farther into the kyanite zone, some generally poikiloblastic garnets have rims devoid of quartzofeldspathic inclusions but with rare kyanite inclusions. This, together with the presence of inclusion-free subhedral garnets in the matrix, indicates the coeval growth of garnet and kyanite. Kyanite preferentially occurs in biotite-rich lithons locally intergrown with biotite laths ((Plate 14)d). These features indicate that reaction 2 was superseded by or accompanied at higher grades by the reaction:

staurolite + muscovite + quartz = garnet + kyanite + biotite + H2O (Dempster, 1985) (Reaction 3)

Locally, kyanite is elongate parallel to S2. However, many kyanites grow across the S2 mica fabric, indicating that they postdate D2. They tend to be smaller and more ragged in form within muscovite-rich bands, consistent with subsequent growth or recrystallisation of muscovite. However, S2 is deflected around some of the kyanites and others are deformed adjacent to garnets wrapped by S2. This relationship, together with the discordance between internal and external fabrics, suggests an overlap with D2. Relationships between kyanite and D3 crenulations are not clear. Kyanites crenulated by D3 have not been recorded here, although Harte and Johnson (1969) reported such textures from Glen Clova. The balance of evidence suggests that kyanite growth occurred during and after D2, but probably before D3.

Sillimanite Zone

The sillimanite zone, marked by the extensive, although not ubiquitous, development of fibrolitic sillimanite, with local prismatic sillimanite, is the most extensively developed of the Barrovian zones in the Ballater district (Figure 17). The rocks are medium to coarse grained, migmatitic and significantly coarser than those in the staurolite and kyanite zones. Many are massive, having recrystallised after D2 (Harte and Johnson, 1969) to the extent that there is no obvious parting. Others, however, preserve S2 crenulated by D3, allowing relationships between deformation and metamorphism to be elucidated. Mineral assemblages are similar to those in the staurolite zone with the addition of kyanite and sillimanite. Staurolite and kyanite generally become less abundant with increasing distance into the zone; kyanite is absent from many rocks in sections in the Waters of Lee and Unich and associated tributaries whereas staurolite occurs principally as inclusions within garnet. In the Queen's Hill Gneiss Formation to the east of Glen Muick, kyanite is restricted to small grains at the cores of fibrolitic masses. Staurolite is generally absent; however ragged grains, remnants from much larger porphyroblasts, occur in a pelitic horizon of unusual composition within the formation [NO 3655 9180]. The host rock contains abundant plagioclase and K-feldspar, together with green biotite, garnet and fibrous sillimanite, but very little quartz. The staurolite contains more than 2 weight per cent ZnO and is cored by the zincian spinel, gahnite. Its survival deep within the sillimanite zone was attributed by Goodman (1993) to the stabilising effect of zinc and low availability of reactant quartz. Small relict grains of zincian staurolite have also been recorded within sillimanite-grade rocks in Glen Doll (Schumacher, 1985). More oxidised rocks here contain relatively simple quartz + plagioclase + biotite + muscovite ± garnet assemblages (see Chinner, 1960). Apatite is a prominent accessory mineral.

Fibrolitic sillimanite forms 20 per cent or more of the modal composition of some rocks. It occurs enclosed within muscovite, biotite and garnet and more rarely as mats within quartz and plagioclase, indicating that it is the product of several different reactions. Some of the most revealing textures as to its origin are found on Ben Reid [NO 32 75]. Fibrolite and prismatic sillimanite occur most commonly in a spatial association with biotite, either as needles within biotite or more abundantly as 'mats' intimately intergrown with biotite. Chinner (1961) ascribed such textures to the preferential nucleation of sillimanite on biotite. The development of this texture involves, initially, the intergrowth of sillimanite and biotite followed by the replacement of biotite by sillimanite with relicts of biotite preserved in the 'mats' of fibrolitic sillimanite. This is interpreted as being due to a complex reaction like:

kyanite + biotite (1) = sillimanite + biotite (2) (Reaction 4)

Plagioclase may also be involved since biotite–sillimanite intergrowths apparently replace plagioclase locally. Small (0.3 mm or less) euhedral garnets occur within fibrolite-biotite mats indicating coeval garnet growth (see below).

Fibrolitic sillimanite has also developed in muscovite porphyroblasts that replace kyanite. Several stages in this replacement are preserved, including the development of ragged kyanite plates enclosed by muscovite and relict 'islands' of kyanite, some in optical continuity, within muscovite. Some such muscovites contain fibrolitic sillimanite, ((Figure 14)e), although fibrolite and kyanite are rarely in contact. This texture may indicate either an indirect polymorphic inversion of kyanite to sillimanite or hydration and retrogression, as seen in the staurolite and kyanite zones, with subsequent development of sillimanite. Plagioclase is corroded and replaced by these muscovites and is also embayed by quartz, with some specimens from Ben Reid showing an intergrowth of quartz and plagioclase.

Farther north, for example east of Allt Cholzie [NO 350 886] and on Creag Mullach [NO 412 952], within highly aluminous pelites of the Queen's Hill Gneiss Formation, the sillimanite forms bladed masses of fibres, their shape indicating that they are replacing kyanite. The presence locally of small kyanite relicts at the cores of the fibrolitic masses ((Figure 14)f) without intervening muscovite confirms that direct polymorphic inversion does occur in places.

South-east of Glen Muick, in the area around and to the south-west of Pannanich Hill [NO 39 93], the Queen's Hill Gneiss Formation contains the assemblage quartz + plagioclase + K-feldspar + biotite + garnet ± sillimanite (Goodman, 1991). The absence of muscovite indicates the reaction:

muscovite + quartz = sillimanite + K-feldspar + H2O (Reaction 5)

Fibrolitic sillimanite mats with rare associated biotite are locally intergrown with, and more rarely overgrown by, muscovite which contains fibrolite. It is not clear whether this indicates contemporaneous growth, and hence similar P–T conditions, for both the development of biotite–sillimanite and muscovite-sillimanite, or whether one is replacing the other.

Subhedral garnets contain inclusion-free rims encircling inclusion-bearing cores. The rims are somewhat pinker in colour than the cores in plane polarised light. Inclusions are mostly coarser grained than in the staurolite and kyanite zones and few preserve any relict fabric. In marginal areas of the sillimanite zone, particularly on Ben Reid, 'atoll' garnets have their cores replaced by biotite, muscovite and fibrolitic sillimanite, whereas the inclusion-free rims are preserved. Several stages in the replacement of the garnet core are recognised. Biotite and muscovite appear to be the first products, as seen in the kyanite zone, followed by the development of fibrolite–biotite intergrowths. Some specimens show fibrolite growth only in the muscovite. The most advanced stages of replacement show mainly fibrolite with only small amounts of interstitial biotite. Sillimanite pseudomorphs of garnet are thought to develop subsequent to garnet replacement in the majority of cases. In several examples, garnet rims are partially or completely isolated from the pseudomorphed cores such that ultimately only strips of garnet occur within the matrix.

Farther north within the sillimanite zone, garnets are commonly large (5 mm or more across), but some specimens contain numerous small (0.5 mm or less across) garnets, with a pale pink colour in plane polarised light, similar to the garnet rims recorded from specimens found on Ben Reid. Some large garnets contain no inclusions; others include staurolite and rarely biotite and staurolite. The latter may comprise up to 20 per cent of the garnet, and commonly occur within a discrete inner zone of the garnet, a feature not seen in the kyanite and staurolite zones where staurolite tends to include garnet (see above). In many parts of the sillimanite zone, staurolite is only preserved as inclusions within garnet. This abundance of included staurolite contrasts with the near absence of early staurolite close to the boundary between the kyanite and sillimanite zones. No obvious explanation can be found other than the occurrence of a higher concentration of early staurolite in these rocks and possible overstepping of reactions due to rapid temperature increase in these rocks. Biotite and muscovite are also found as inclusions within garnet, together with quartz, plagioclase and opaque minerals. These features indicate that significant garnet growth was associated with the sillimanite zone although many specimens show the development of sillimanite without garnet. Many garnets overgrow or include sillimanite ± biotite intergrowths, in contrast to 'atoll' garnets which are replaced by sillimanite ± biotite. The textural evidence indicates that garnet growth occurred as a result of the breakdown of remaining staurolite by the reaction:

staurolite + muscovite + quartz = garnet + sillimanite + biotite + H2O (McLellan, 1985) (Reaction 6)

Many garnets which contain staurolite inclusions are partly wrapped by biotite–sillimanite intergrowths with no adjacent muscovite, whereas other garnets enclosed within both muscovite and biotite–sillimanite inter-growths contain no staurolite or quartz inclusions. Such textures indicate that the reaction locally proceeded until one of the reactants was consumed. Some specimens do not show such complex textures and only contain fibrolitic sillimanite within muscovite porphyroblasts which overgrow the regional schistosity.

In these rocks sillimanite is generally developed within S2 although it is not confined to the schistosity plane. It also occurs within post- D2 muscovites which grow across S2 and replace syn- to post- D2 kyanite, indicating that sillimanite growth postdates D2. Sillimanite is generally deformed by D3 crenulations. However, several examples of sillimanite developing along, and more rarely overgrowing, D3 fold axial surfaces are seen ((Plate 14)g). In fact sillimanite cross-cuts crenulated fibrolitic sillimanite ((Plate 14)h). Rare sillimanite within post- D3 muscovites has also been noted. These relations indicate that while most sillimanite growth occurred between D2 and D3, there was some overlap with D3. Sillimanite growth apparently overlaps the development of high strain fabrics in rocks close to the boundary between the Tarfside Psammite Formation and the Longshank Gneiss Formation. Needles of sillimanite intergrown with biotite together with elongate magnetite define a strong fabric and associated coaxial tight crenulation in a magnetitetourmaline-apatite-biotite-sillimanite-garnet rock. Many sillimanites show basal sections in the hinges of the crenulations, whereas others grow across the strong fabrics and the crenulation. A similar strong fabric in a specimen from an adjacent locality has a strong shear fabric defined by quartz ribbons with associated asymmetric tails in porphyroclasts. Since sillimanite postdates D2 and the shear fabric is deformed by D3 structures, shearing must have occurred between D2 and D3.

Locally, garnets are separated from sillimanite–biotite intergrowths by a rim of muscovite with or without quartz, the quartz locally protruding into the garnet. Some of these muscovites contain fine-grained staurolites with well-formed crystal faces. This association is interpreted as the product of a reversal of the garnet + sillimanite-forming reaction:

garnet + biotite + sillimanite = staurolite + muscovite + quartz (Reaction 7)

Late staurolite growth is prominent in thin sections from Ben Reid. Parts of the isolated rims of 'atoll' garnets and some relicts of garnet core are enclosed within staurolite. These staurolites are up to 2 mm across, preserve straight crystal faces and contains few inclusions, in contrast to both relict staurolites within the matrix and inclusion-rich staurolites in the kyanite and staurolite zones. Similar late staurolites occur within fibrolitic sillimanite mats; again they display straight crystal faces, oblique to the fibrolite mats. However, where they are in contact with muscovite, grain boundaries of these staurolites are somewhat embayed. Further evidence of the late growth of staurolite is seen where they partially include kyanite; several such kyanites contain biotite inclusions. Although grain boundary relationships are in part ambiguous, several of these staurolites display straight crystal faces suggesting that the staurolite also developed by reversal of the kyanite-producing reactions, that is:

garnet + kyanite + biotite = staurolite + muscovite + quartz (Reaction 8)

In contrast to the ragged kyanite cores described above, small (0.2 mm), near euhedral crystals occur within mats of fibrolitic sillimanite in a specimen from Ben Reid ((Plate 14)f). This is one of the few occurrences of kyanite and sillimanite in contact in the southern part of the sillimanite zone. The texture is indicative of kyanite replacing sillimanite. Similar near-euhedral kyanites occur within muscovite contrasting with the relict kyanite preserved in muscovite in the same rocks. Textural relationships between the late kyanite and staurolite are not seen. Similar late growth of kyanite and staurolite has been described by Chinner (1961). McLellan (1985) only recognised the late growth of staurolite which she attributed to the reaction:

garnet + sillimanite + H2O = staurolite + quartz (Reaction 9)

Again, this is the reversal of one of the potential sillimanite-forming reactions. Late staurolite growth may be represented at lower grades by the inclusion-free staurolites, and kyanite overgrowths in the kyanite zone may also have developed at this time.

Retrogressive metamorphism

Evidence of retrogressive metamorphism occurs only patchily throughout the district with the result that in areas such as Glen Clova the complex prograde metamorphic textures are largely preserved. In those metasedimentary rocks which are retrogressed biotite, garnet, diopside and hornblende become altered to chlorite, while both plagioclase and alkali feldspar react to form fine white micas. In the headwaters of Glen Esk, late chloritisation of garnet and biotite is accompanied by replacement of sillimanite by fine-grained white mica, although as in Glen Clova many rocks in this area have also largely avoided retrogression. Over much of the sillimanite zone sillimanite is unaffected by retrogression, and swirls and cords of fibrolite survive in some heavily retrogressed samples. In addition to the chloritic alteration of hornblende, biotite, garnet and clinopyroxene in calcsilicate-rocks, these mineralogies also show widespread alteration to a fine-grained clay-like material that is almost opaque in thin section.

Retrogression is less pronounced in the metabasic rocks; the principal effect is the replacement of hornblende by biotite, which in turn may be retrogressed to chlorite. Diopside and garnet are also replaced by chlorite, and feldspars alter to fine white micas.

Migmatisation

Under the effects of high-grade Barrovian metamorphism most of the Dalradian rocks, especially semipelitic and pelitic lithologies, have formed gneisses typically with a strongly banded appearance as the mafic and quartzofeldspathic parts became differentiated. At higher metamorphic grades so the distinction between the layers and the volume of the quartzofeldspathic bands (leucosomes) increases and the rocks take on the typical mixed appearance of migmatites.

Two main episodes of migmatisation are recognised in the Ballater district. The earlier episode occurred throughout the sillimanite zone rocks and in part of the kyanite zone, although the limit of migmatisation with respect to the kyanite zone cannot be clearly established because of poor exposure. Early formed migmatites within the metasedimentary rocks are typically stromatic, with planar bands and stringers of leucosome. Although the leucosome can be locally very abundant, metamorphic segregation, rather than melting seems to be the dominant mode of formation. The metabasic rocks may also be stromatic, but more commonly they are agmatitic, with an irregular network of leucosome veinlets surrounding blocks of metabasic material.

The leucosomes are of leucotonalitic composition, typically less than 2 cm thick, and composed largely of quartz and plagioclase with minor muscovite. The plagioclase ranges from oligoclase in the south-east to andesine in the north, and is similar to that in the melanosome. Only migmatites of talc-silicate composition contain K-feldspar in the leucosome, presumably due to the abundance of K-feldspar in the bulk of the calcsilicaterock. Elsewhere K-feldspar is absent from the migmatitic leucosome, precluding partial melting as a mode of formation (Yardley, 1978).

Leucosomes define F3 folds and therefore predate D3. They are apparently concordant with the regional S1/ S2 fabric and in places they define tight to isoclinal folds. Structural relationships with D2 are ambiguous, although the migmatisation is considered to broadly relate to D2.

Leucosomes are strongly attenuated with the development of rodding lineations in places, particularly in the Burn of Doune in Glen Mark. It is not clear whether this is the result of D2 deformation or shearing between D2 and D3.

Migmatites of the second generation have a more restricted distribution, occurring only sporadically within the sillimanite zone, but locally are more intensely developed than the first generation. The main outcrops, shown in (Figure 17), include most of the areas mapped by Barrow (1912) as oligoclase-biotite gneiss. Those in the north-east part of district lie within the Queen's Hill Gneiss Formation. The larger of these, encompassing the occurrences west of Craig Vallich, is known as the Pannanich Hill Complex and was described in detail by Goodman (1991). In addition there are smaller developments south-east of Drum Cholzie [NO 35 87], to the east of the Coyles of Muick [NO 33 91] and [NO 34 91] and on the south side of Craig Megen [NO 31 89]. In the other main locus of migmatisation, between Glen Clova and Glen Mark, the rocks lie within the Longshank Gneiss Formation.

In the second generation migmatites the leucosomes are coarse grained and locally pegmatitic, occurring in the form of lenses and nebulous patches. With increasing segregation, the rocks grade into schlieric migmatites in which relict mica-rich laminae are progressively enveloped. Ultimately, the migmatites are completely reconstituted to produce massive medium- to coarse-grained tonalitic rocks of igneous aspect, with or without mafic schlieren. The bulk of the reconstituted rock (neosome) is formed of a coarse-grained mosaic of biotite, oligoclase and quartz, with or without muscovite, with local mafic clots of garnets rimmed, and in part replaced, by biotite. Migmatites in the Queen's Hill Gneiss Formation on the north-west side of Glen Muick for example [NO 3299 8931] additionally contain prismatic sillimanite.

In the Pannanich Hill Complex, the neosomes contain small restite inclusions representative of the common lithologies of the Queen's Hill Gneiss Formation. Those of semipelitic and pelitic type grade into the surrounding oligoclase-biotite gneiss without any obvious reaction rim, suggesting again that these lithologies have acted as a protolith for the gneiss. Reaction textures within the pelitic inclusions indicate reaction between garnet, biotite and sillimanite to form a symplectite of hercynite and cordierite. This reaction is similar to that seen at the highest grades of thermal metamorphism in the aureole of the Caledonian granitoids (Chapter 14), though it should be noted that unlike the later hornfelses, the pelitic inclusions in the oligoclase-biotite gneiss contain no thermal andalusite.

Inclusions of psammitic (Plate 12) or calcsilicate metasedimentary rocks have distinct reaction rims, indicating that they were not in equilibrium with the matrix. Calcsilicate lithologies form a higher proportion of the inclusions than is apparent in the bulk of the Queen's Hill Gneiss Formation, emphasising the refractory nature of this material. Metabasic inclusions, from thin bands in the formation, also have broad reaction rims, which are generally biotitic. Within these inclusions, reaction has taken place between hornblende, garnet and clinopyroxene to produce fine reaction fringes of an anthophyllite ilmenite-feldspar symplectite. Trails of amphibolite inclusions, which trace out the location of the original metabasic bands along strike, can be followed for several hundred metres along Pannanich Hill spur, indicating in situ generation of the surrounding migmatites. Apart from the semipelitic to pelitic inclusions, which have wispy and indeterminate shape, the inclusions are rounded or oval, due to corrosion and reaction with the matrix.

The nebulous and schlieric migmatites cross-cut the stromatic migmatites wherever relationships are preserved. This is particularly well displayed in Glen Mark where the younger leucosomes overprint the older strongly deformed stromatic leucosomes, indicating that they postdate D2. Elsewhere, it is commonly difficult to distinguish the two periods of migmatisation, either because early deformation is less intense or because the stromatic migmatites recrystallised during the younger event. Relationships farther west in the Glenshee district (Sheet 56W) clearly show that the younger period of migmatisation predates D3 (Crane, et al., in press). McLellan (1989) also recognised two periods of migmatisation in the eastern Grampian Highlands similar in aspect to those described here. She attributed the early stromatic leucosomes to metamorphic segregation and the later schlieric pervasive leucosomes to partial melting. However, no evidence to substantiate partial melting has been found during the present study.

A third period of leucosome segregation related to D3 deformation is demonstrated by pegmatitic segregations along the axial surfaces of F3 folds within the Rough Craig Granite (Robertson, 1991; fig 45). Farther north for example at [NO 3602 9090] both metasedimentary and metabasitic rocks contain patches of comparable pegmatitic leucosome, which not only cross-cut or obliterate earlier migmatisation, but also demonstrably postdate the F3 folds. The diffuse irregular shape of the migmatites indicates that they probably formed when no deformation was taking place.

Although direct evidence of the relationship between metamorphic mineral growth and migmatisation is lacking because porphyroblasts do not occur in the leucosomes, timing relative to the structural history suggests that the migmatitic history apparently reflects closely the metamorphic history. The formation of stromatic leucosomes was apparently coeval with the development of the staurolite–kyanite grade metamorphism, the schlieric diatexites with the sillimanite grade metamorphism and the segregations along the axes of the F3 fold structures possibly reflecting the renewed garnet–staurolite–kyanite grade metamorphism.

Implications and inferences of textural evidence

All recent studies indicate that the main metamorphic porphyroblast growth occurred between D2 and D3 with some overlap with D3. However, there has been some disagreement over the significance and timing of sillimanite growth. Chinner (1966) and Harte and Johnson (1969) considered that sillimanite overprinted the Barrovian zonal sequence and was the product of a separate thermal pattern. Harte and Hudson (1979) and McLellan (1985), however, related sillimanite growth to a prograde Barrovian metamorphism and thus saw no need for a subsequent sillimanite overprint. Dempster (1985) related sillimanite growth to the uplift phase of a single thermal event.

Observations made during study of the Ballater district indicate that metamorphism overlapped both the D2 and D3 tectonic events. The presence of apparently pretectonic porphyroblasts may reflect the nature of D2 with strain partitioned into discrete domains so that, while all porphyroblasts may have grown during the regional D2 deformation event, some grew in actively deforming domains and others grew in static domains that were subsequently deformed. In the higher grade rocks peak metamorphism occurred later with respect to the overall tectonic history. In the staurolite zone, it was approximately coeval with D2 as shown by the growth of poikiloblastic garnet and staurolite. In the sillimanite zone, the textural relationships of sillimanite indicate that the peak was pre- to syn-D3 .

Metamorphism can be broadly divided into four phases. The first (M1) is marked by the occurrence of relicts of garnet and staurolite within biotite porphyroblasts. Little is known of the textures or relationship of this period of mineral growth to the remainder of the metamorphic history. The second phase (M2) led to the progressive development of garnet, staurolite and kyanite in appropriate zonal areas. This was followed by the widespread development of muscovite and formation of local 'atoll' garnets. The third phase (M3) was marked by the development of sillimanite with garnet and biotite. Minor renewed growth of staurolite and local kyanite postdated sillimanite and constitutes the fourth phase of metamorphism (M4). There is no evidence for any significant time gaps between each of these phases and therefore metamorphism may be viewed as a single progressive event spanning D2 and D3. Metamorphic zones are the composite product of these phases and therefore only give information on the peak assemblages preserved within any particular area rather than yielding regional metamorphic conditions at a particular point in time. This is particularly apparent with respect to the sillimanite zone which reflects post- D2 to pre- to syn- D3 metamorphic conditions. It is separated in time from the largely D2 staurolite and kyanite zones by a period of hydration and may be related to a different thermal pattern. Similarly, late staurolite and local kyanite represent a further change in metamorphic conditions.

P–T estimates have been interpreted to indicate that the Barrovian zones were the product of high lateral thermal gradients since they show little relationship to depth of burial (Harte and Hudson, 1979; Dempster, 1985). Such thermal gradients were postulated to occur adjacent to a major tectonic discontinuity close to the present Highland Boundary Fault (Harte and Hudson, 1979) which had a lateral chilling effect on the eastern Scottish Dalradian high-grade assemblages (Harte and Dempster, 1987); the metamorphism was related to the widespread intrusion of granitic magma at depth (Harte and Hudson, 1979). Chinner (1980) recognised an inverted thermal gradient in the Glen Clova–Glen Esk region; it was suggested that this represented the southeast limb of a recumbent thermal anticline. This thermal anticline was postulated to extend north-eastwards from Ben Vuirich through the Duchray Hill Gneiss Member, along the zone of Barrovian kyanite and sillimanite grade metamorphism, and was considered to represent a synmetamorphic thermal disturbance, possibly 'due to the introduction of warm material with the upward and southward translation of the Tay Nappe' (Chinner, 1980).

Metamorphic zones within the Glen Clova–Glen Esk area appear to be broadly parallel to major lithological boundaries and folded by D3 structures (Figure 17), consistent with the textural evidence of porphyroblast growth before D3. Metamorphic grade and grain size increase northwards, such that stratigraphically older rocks are preserved at higher grade and are generally coarser grained than lower grade equivalents. To the south of the Ballater district the zone boundaries are thought to dip to the south-east (Harte and Hudson, 1979). There is insufficient evidence in the district to determine the three-dimensional orientation of the staurolite-kyanite boundary. The kyanite–sillimanite zone boundary is rather better constrained. Here the disposition of this boundary indicates that the metamorphic gradient was the right way up, contrary to what was reported by Chinner (1980), except where it was inverted by F3 folds.

Possible qualitative P–T-time paths are shown on a simple petrogenetic grid in (Figure 18). Textures associated with M2 metamorphism are consistent with prograde metamorphism in which higher grade assemblages occur structurally later and farther north than lower grade assemblages. The incoming of muscovite and development of 'atoll' garnets may represent a period of hydration of M2 assemblages and, although the associated P–T conditions are not known, it is assumed to represent a period of retrogression in (Figure 18). The M3 phase represents a substantial rise in temperature into the sillimanite field. The M4 phase suggests a return to M2 conditions. The M3 thermal perturbation is broadly coeval with metamorphic grain coarsening, and the development of pervasive migmatites and reconstituted rocks as seen in the Duchray Hill Gneiss Member, the Queen's Hill Gneiss Formation and parts of the Longshank Gneiss Formation (see below). Goodman (1991) attributed metamorphic reconstitution in the Queen's Hill Gneiss Formation to the combined effects of enhanced heat flow from apparently penecontemporaneous 'younger basic' intrusions and shearing with associated movement of fluids. A similar origin for the M3 thermal perturbation could be invoked. Support for this hypothesis comes from the regional distribution of sillimanite-bearing rocks and the basic intrusions of the north-east part of the Dalradian outcrop.

Geothermobarometry

Previous estimates of Barrovian metamorphic conditions in the south-east Highlands yield temperatures ranging from about 550°C in the staurolite zone to about 650°C in the sillimanite zone (Harte and Hudson, 1979; McLellan, 1984; Dempster, 1985) to 820°C in the area south-east of Glen Muick (Baker and Droop, 1983). Metamorphism was interpreted as near isobaric with pressure about 6 kb, implying a high lateral thermal gradient. The staurolite zone spans a temperature range of only 20 to 30°C at 6 kb and the kyanite zone about 50°C using the Al2SiO5 system of Richardson et al. (1969) and the petrogenetic grid of Harte and Hudson (1979). The kyanite zone is reduced to a temperature range of less than 20°C using the Al2SiO5 system of Holdaway (1971). The restricted occurrence of the kyanite zone in Glen Clova is compatible with such a narrow temperature range.

During this study, P–T estimates have been calculated from four samples. These comprise one from the staurolite zone east of Milton of Clova (S79402), two from the margin of the sillimanite zone on Ben Reid ((S79416) and (S79417)) and one from several kilometres into the sillimanite zone in the Water of Unich (S79470). Temperature estimates are based on the Ferry and Spear (1978) garnet–biotite geothermometers; the more recent Hodges and Spear (1982) geothermometer, which takes account of impurities in the garnet, is considered by many workers to yield unrealistically high temperatures. Pressures are based on the Newton and Hasleton (1981) garnet–plagioclase geobarometer. Biotites show only a narrow range in composition within any particular specimen. Garnets, however, show considerable variation and thus when combined with biotite yield a range of P–T values. In general, points about 50 microns from the garnet margin appear to give results most compatible with previous studies. Such features, when combined with the heterogeneous metamorphic textures already outlined, create problems in selecting and analysing equilibrated mineral pairs, which are an essential requirement for reproducible geothermobarometry. On this basis the data presented here should be used with caution.

Specimen (S79402) from the staurolite zone yields P–T values of 580°C and 5.8 kb (Figure 19). Temperatures are based on analysing garnet points 50 microns from garnet rims and matrix biotites, and pressures using the same garnet points with plagioclase inclusions in the garnet. The results are consistent with other P–T estimates from the staurolite zone. Calculated temperatures and pressures are significantly lower if points 15 microns from garnet rims are used in the calculations.

Specimen (S79416) from the margin of the sillimanite zone on Ben Reid, yields P–T estimates of 630°C and 6 kb (Figure 19) from the rims of 'atoll' garnets, biotites included within or adjacent to garnets and plagioclase at garnet margins. Slightly lower temperatures (about 600°C) are obtained by analysing the relicts of garnet cores. Since specimen (S79470) does not contain plagioclase, pressure could not be calculated. However, assuming a pressure of 6 kb, estimated temperatures of about 630°C are obtained by analysing a range of positions at either garnet rims or 250 microns into garnets and biotites from the matrix or in contact with garnet. However, one garnet–biotite pair yields a temperature of about 680°C. The results from the sillimanite zone are generally in agreement with previous P–T estimates of about 650°C and 6 kb (Harte and Hudson, 1979; Dempster, 1985; McLellan, 1985).

Buchan metamorphism

North and west of Glen Girnock the high incidence of large Caledonian granitoid bodies has resulted in widespread contact metamorphism in the Dalradian rocks such that the regional metamorphic textures and mineral assemblages have been widely overprinted. However, most thin sections preserve part of the regional assemblage composed of quartz, plagioclase, and brown biotite partly replaced by chlorite. The well-developed preferred orientation of biotite and chlorite defines a strong tectonic fabric (S2). The rocks also include relicts of garnet porphyroblasts which apparently developed either pre- or syn-D2, as indicated by quartz pressure shadows preserved in several examples. Staurolite was noted only in one specimen (S82068) where it occurs as a relict grain within andalusite, which, in turn, is mantled by cordierite and biotite.

Sillimanite has more widespread occurrence, but it appears to have developed during both regional and thermal events. For example, in (S77613), from [NO 2394 9552] elongate prismatic crystals are aligned parallel to the foliation and would seem to be regional in origin. They occur within quartz and K-feldspar grains and are truncated by thermal andalusite ((Plate 15)a).

Chinner and Heseltine (1979) recorded regional andalusite at two localities just north of Glen Girnock. Possible regional andalusite ((Plate 15)b) was recorded during this survey in an isolated exposure of aluminous semipelite (S77613), 200 m west-south-west of Drumargettie [NO 2414 9558].

The thermal overprint, although never totally absent, is much reduced on the south-east side of Glen Girnock such that the pelitic and semipelitic rocks of the Creag Liath Pelite and Knock Semipelite members provide evidence of regional conditions. Both members are characteristically rich in muscovite; additionally they contain quartz, biotite, plagioclase, garnet, staurolite, cordierite and andalusite

((Plate 15)c), some of which are attributable to regional metamorphic events. Sillimanite and corundum were also recorded. The rocks have a penetrative schistosity defined by mica and elongated quartz and feldspar. The fabric is equated with the regional S2; in places it is folded by D3 structures and an incipient S3 crenulation cleavage is developed.

A striking aspect of several thin sections is the garnet and staurolite porphyroblasts, particularly the latter which may be up to 5 mm long. The garnets, which are susceptible to replacement by biotite, commonly contain inclusion trails of quartz, plagioclase and iron oxide, typically at a high angle to the foliation and more rarely displaying sigmoidal patterns. The additional presence of pressure shadows suggests that the garnet formation is pre- D2 to syn-D2 . Staurolite crystals and their inclusion trails (quartz or iron ore) typically align themselves with the schistosity, although S-shaped and high-angle inclusion trails and pressure shadows are also present. The evidence suggests that staurolite growth largely postdated that of garnet, and in regional chronological terms was syn-D2 .

Sillimanite is relatively uncommon in these rocks but it occurs both in fibrolitic and prismatic habits. Where prismatic, it generally occurs within biotite, reminiscent of textures seen in the adjacent Queen's Hill Gneiss Formation. Although these rocks should lie within the limits of Barrow's kyanite 'wedge' no trace of the mineral was recorded during this survey, or by Tadesse (1991) or Chinner and Heseltine (1979).

Implications of textural and petrological evidence

In spite of the thermal overprint it is possible to distinguish two Buchan type metamorphic zones in the area north-west of the Muick–Girnock watershed. The more north-westerly is characterised by the assemblage andalusite + garnet + sillimanite whereas that to the south-east comprises staurolite + garnet + andalusite + sillimanite. Locating the position of the interzonal boundary is hindered by a lack of suitably diagnostic lithologies (the area being dominated by calcsilicate- and amphibolitic rocks), but Tadesse (1991) found sufficient evidence to trace out a staurolite-in isograd (Figure 17).

Comparable staurolite-rich rocks occur on the Banffshire coast between Boyndie Bay and Whyntie Head where they are considered to be part of a transition towards higher pressure, Barrovian type metamorphism (Hudson, 1980).

The position of the andalusite–kyanite boundary is constrained by the occurrence of muscovite-rich rocks with staurolite but no kyanite at [NO 317 917] and the presence of relict kyanite partly replaced by sillimanite within the Queen's Hill Gneiss Formation (see also Chinner and Heseltine, 1979), respectively, immediately north-west and south-east of the Coyles and Muick Shear Zone. The andalusite–kyanite boundary, and hence the Barrovian–Buchan distinction is thus placed at the Shear Zone.

The Buchan metamorphic chronology is broadly comparable with that described for the Barrovian rocks, although events corresponding to M1 and M4 have not been recognised. The principal andalusite and garnet growth would appear to correspond to M2, with staurolite and muscovite developing during the post M2 hydration event. As in the Barrovian rocks sillimanite formed during Ms.

P–T estimates for the Buchan type staurolite zone are 510 to 550°C and 2 to 3.5 kbar, whereas comparable figures for the andalusite zone are 490 to 510°C and 1.7 to 2 kbar (Hudson, 1985). The difference in P–T conditions implied for the Buchan-type staurolite zone and Barrovian kyanite zone supports the concept of a significant tectonic break across the Coyles of Muick shear zone. Main movement on the shear zone can be assigned to syn- D2 to late-D2, being effectively bracketed by the early growth of staurolite and kyanite and the late development of sillimanite.

Chapter 12 Post-tectonic major intrusions—introduction and diorite–granodiorite

Introduction

The waning stages of the Caledonian Orogeny in northeast Scotland were characterised by the addition to the crust during the late Silurian of copious volumes of granitoid magma, which gave rise to the so-called 'Newer Granites' of Read (1961). There is a notable concentration of these intrusive rocks in a 30 km-wide east–west zone straddling the Dee valley and extending for over 100 km from the Cairngorm and Monadliath granites to the east coast. The intrusions are roughly coeval and are regarded as protrusions from the East Grampians Batholith (Plant et al., 1990), which gravity data indicates underlies much of the eastern and central Highlands (Rollin, 1984). The zone is dominated by the large pink granites of the Cairngorm Suite of Stephens and Halliday (1984), typified by the Cairngorm, Ballater and Mount Battock plutons. There is, however, a range of compositions present, encompassing granite–granodiorite complexes (for example Crathes, Torphins (Gould, 1997)) and less common and smaller diorite–granodiorite plutons and complexes (for example Glen Doll, Abergeldie Complex); this last group shares characteristics with the South of Scotland Suite of intrusions (Stephens and Halliday, 1984).

The Ballater district lies on the southern edge of the zone and contains representatives of both pink granite and diorite–granodiorite groups, which crop out over an area of about 200 km2 (Figure 13). The diorite–granodiorite group, described in this chapter, is part of a suite of small scattered bodies which is confined to a 35 km long and 15 km wide north–south area extending from Loch Kander (Sheet 65W) to the east end of the Cairngorms and from Glen Doll to Gairnshiel (Sheet 75E). The area is also occupied by parts of the Cairngorm, Glen Gairn and Lochnagar major granitic plutons as well as many smaller granitic bodies. The more evolved members of the diorite–granodiorite group are concentrated approximately in the centre of the area, forming the small Abergeldie Complex of intrusions, which range from diorite to white granite and which may have compositional links with the granites of Lochnagar and Glengairn.

The Lochnagar and Glen Gairn granites occur as two large plutons and a number of satellite bodies, and include both grey-pink magnetic monzogranites and evolved pink granites. The other major plutons present, Ballater and Mount Battock are somewhat different; they are distinct compositionally, there are no mafic phases associated with them and the granites are all pinkish and poorly magnetic at outcrop. The Coilacriech Granite is similar to the Ballater granite, but is finer grained. All these granite bodies are described in Chapter 13. Petrological descriptions are based on Phillips (1991).

Diorite–Granodiorite Group

The diorite–granodiorite bodies marginal to the southern and eastern sides of the Lochnagar granites were regarded by Oldershaw (1974) as remnants of a 15 km diameter pluton which formed the first intrusive phase of the Lochnagar Granite. However, these bodies do not, as he indicated, form a continuous rim around the eastern half of the granite, but form a series of separate and, for the most part, petrographically different bodies. Like numerous small bodies in the north of the Ballater district, in the Braemar district (Sheet 65W) and farther north, they are now regarded as a series of separate intrusions consanguineous with, but earlier than, the Lochnagar and Glengairn granites. This essentially reverts to the original interpretation of these bodies by Barrow and Cunningham Craig (1912), who recorded that the bodies north of the Dee ranged in composition from augite-bearing diorite to biotitegranite. It is the concentration of this range of compositions in the area between the Lochnagar and Glengairn granites which will be described as the Abergeldie Complex. Additionally, in the Ballater district there are five bodies predominantly composed of diorite, named the Glen Doll, Moulzie Burn, Juanjorge, Allt Darrarie, and Cul nan Gad diorites. The last three of these lie marginal to the Lochnagar Granite and are veined, and are locally foliated, by the earliest phase of the Lochnagar granites.

Glen Doll Diorite (HD)

The outcrop of the Glen Doll Diorite pluton, which occupies nearly 12 km2, is crudely oval shaped with a longer north-east-trending axis, and sits astride the Glen Doll Fault. The diorite is confined mostly to ground below 600 m, the upper slopes comprising upper Argyll and Southern Highland Group metasedimentary rocks on the roof and sides of the intrusion. Exposure is not plentiful on the lower slopes, which are extensively forested, and bedrock in the valley bottoms is almost completely obscured by alluvium. This contrasts with the well exposed upper slopes, but even there field relations are partly obscured by landslip, most notably on Craig Mellon and Cairn Broadlands.

A detailed description of the Glen Doll Diorite is contained in Jarvis (1987), from which the following account is largely drawn. In summary, the pluton is dominated by hornblende and pyroxene diorites, the remainder comprising gabbro, monzonite, monzogranite and granite. The diorites are predominantly medium grained, though fine and coarse varieties are also present, and are markedly heterogeneous and xenolithic.

The diorite suite includes relatively anhydrous varieties bearing significant clinopyroxenes and orthopyroxenes. With plagioclase in the range An49–62 some of these rocks are closer to gabbro than diorite. Indeed Mahmood (1986) recorded gabbro at a number of sites in and around White Water. However, 90 per cent are pyroxenefree or have the pyroxenes restricted to relict cores within the all-pervasive green and brown hornblende. Both pyroxene- and hornblende-bearing diorites contain plagioclase, biotite, apatite and iron oxide with or without sphene and quartz. North and north-east of Braedownie [NO 287 757] disseminated pyrite is widespread with locally developed chalcopyrite. The diorites display igneous layering (interpreted as floor-parallel) and contain cumulate magnetite and ilmenite.

Biotite and green hornblende are particularly common in the south-east part of the pluton where, in places [for example 2918 7590] the diorite encloses patches of appinite with 4 mm-long hornblende megacrysts. With increasing abundance of interstitial to poikilitic perthitic orthoclase, diorite near the south-east edge grades laterally and vertically into quartzmonzonite. The latter is locally heterogeneous and highly xenolithic and contains about 5 per cent biotite in the form of plates, up to 5 mm across.

There is conflicting evidence regarding the nature and extent of the most basic elements of the pluton. Barrow and Cunningham Craig (1912) and Mahmood (1986) considered them to be ultrabasic, respectively describing them as serpentine and picrite and olivine-hornblendepyroxenite, whereas Jarvis (1987) on the evidence of a more detailed sampling programme classified them as gabbro and diorite. Barrow and Cunningham Craig (1912) estimated that they formed up to one fifth of the pluton, although on 1:63 360 scale Sheet 65 (1904) they are shown as occupying less than 5 per cent. Jarvis (1987) stated that they occupy less than 10 per cent of the pluton, being confined to small areas around Moulzie and Kilbo Burns, at the north-east and south-west extremities respectively of the body. The Moulzie Burn occurrence takes the form of a north-easterly tapering wedge between the edge of the pluton and the Glen Doll Fault. The western edge of the gabbro is believed to be defined by the Ketchie Burn Fault, although exposure is poor in that area. The occurrence (Jarvis, 1987) is dominated by medium- to coarse-grained gabbro containing clinopyroxene, orthopyroxene and hornblende with or without biotite. Layering within the gabbro, defined by local differences in grain size and mineralogy, strikes north-west paralleling the eastern edge of the pluton. Appinitic hornblende meladiorite occurs within the gabbro; the contact between the two rocks is sharp but uneven, with lobes of appinite in places protruding into the gabbro. The contact of the gabbro with the hornblende diorite to the south is poorly exposed, but appears to be steep. Both rocks become progressively more fractured towards the contact, which is assumed to be a faulted one.

The form of the Kilbo Burn occurrence is less well defined, for although there is a reasonably good section in the stream bed between [NO 2682 7557] and [NO 2689 7502], exposure in the surrounding forested area is extremely limited. It comprises (Jarvis, 1987) a series of layered cumulate rocks including olivine gabbro, olivine norite, norite, gabbro and monzonite. The layering, which is vertical and parallels the contact with the country rocks, consists of texturally and mineralogically distinctive units between 10 and 100 m in thickness, some of which are repeated, through the sequence.

Jarvis (1987) recorded medium-grained xenolithic monzogranite as a marginal facies along much of the eastern and southern boundaries of the diorite except in the area of Red Craig. The outcrop of the monzogranite, which ranges in width from 20 m in the north-east to 350 m around Kilbo Burn, would seem to be more discontinuous than as shown by Jarvis (1987, fig 3.1). This is well displayed on The Scorrie [NO 27 75] where, on the upper slopes it comprises three discrete bodies up to 50 m wide situated between the country rock and a roof pendant complex of metasedimentary rock and diorite; on the lower slopes [NO 2805 7502] it is restricted to a few thin veins of pegmatite cutting the country rock. The general form of the monzogranite bodies is of vertical cylinders. There is no evidence of chilling between acid and intermediate rocks, but at Capel Road [NO 2925 7837] the monzogranite is clearly seen veining fractures in the hornblende diorite.

The largest area of homogeneous granite recorded during this survey occurs in a 150 m-long section beside a track, 200 m north-east of Red Craig around [NO 2967 7620]. The granite is pink to red, medium to coarse grained, rather weathered and carries some disseminated pyrite.

Contact relationships

The outer margin of the intrusion is well exposed on The Scorrie [NO 280 750] and on the southern face of Craig Mellon around [NO 265 764]. In most places contact with the metasedimentary envelope is near-vertical and sharp. However, on The Scorrie and near Red Craig the contact zone comprises an extremely heterogeneous assemblage of igneous rocks containing xenoliths and screens of high-grade metasedimentary hornfels. On The Scorrie this takes the form of a 500 X 400 m, northerly tapering, wedge-shaped roof to the diorite up to 250 m thick. It is dominated by hornfelsed gneissose semipelite with the assemblage cordierite, plagioclase, magnetite, hercynite and sillimanite. The contact zone includes sporadic small bosses of diorite; elsewhere the metasedimentary rock is intensely net-veined by quartz-diorite.

In the vicinity of Red Craig [NO 295 760] the contact zone is up to 500 m wide. Igneous rocks in the western part of this zone comprise mostly quartz-diorite and quartz-monzodiorite, whereas quartz-monzodiorite, granodiorite and granite form the eastern part. The evidence from Red Craig indicates that most xenoliths are compatible with derivation from close to the present erosion level. However, some material, for example a metacarbonate xenolith, has probably been brought to the present erosion level from deeper levels in the intrusion. None is a roof pendant. The proportion of metasedimentary xenoliths and rafts broadly increases eastwards; immediately west of Red Craig around [NO 295 760] one such inclusion covers an area of 250 by 200 m. Contact with the enclosing quartz-monzodiorite is well exposed at several points. Quartz-monzodiorite sheets, typically about 0.5 m thick, and broadly concordant with a litho-logical layering in the hornfels, occur in the marginal parts of the raft. Some of these sheets show transitional contacts with the hornfels, as zones (about 10 cm wide) of partially assimilated hornfels. These zones are composed largely of perthitic orthoclase, plagioclase and biotite with some quartz (S83481). Orthoclase forms mostly large poikilitic plates which enclose biotite, aggregates of spine' and more rarely cordierite. This lithology grades into metasedimentary hornfels with increasing abundance of cordierite, spinel and sillimanite, and decreasing biotite and plagioclase. Similar textures are seen mantling xenoliths and schlieren within the monzodioritic and granitic rocks. Elsewhere, the contact between the raft and the quartz-monzodiorite is sharp although commonly not planar. A 2 m-wide zone of finer grained quartz-monzodiorite [NO 2937 7604] may represent a chilled margin. However, 200 m east [NO 2958 7601] no such finer grained marginal facies is present, although a 2 cm-wide marginal zone contains biotite plates aligned parallel to the contact.

The eastern part of the contact zone of the intrusion extending north-north-east through Red Craig is more heterogeneous and composed of quartz-monzodiorite, granodiorite, granite and hornfels. The hornfels occurs as xenoliths and mafic schlieren ranging from a few millimetres to 250 m across. Most are psammite or semipelite, although at [NO 2922 7584] a xenolith, 1 m across, is composed of metacarbonate rock. The latter rock has sharp contacts with the encircling quartz-monzodiorite, although relationships are confused by very close jointing with numerous slickensides and areas of broken monzodiorite and felsite.

Contacts between hornfelses and the igneous rocks take a range of forms and many are transitional. In the cliffs south of Red Craig [NO 294 757] clean exposures show a transition from hornfels, through hornfels with veins of intrusive rock, to detached but unrotated hornfels xenoliths, and ultimately to texturally heterogeneous granodiorite or quartz-monzodiorite choked with randomly orientated, millimetre- to centimetre-scale xenoliths and schlieren showing various stages of assimilation. Elsewhere, psammites are preserved as discrete xenoliths whereas originally interlayered semipelites have been largely assimilated and are recognised only as dismembered mafic schlieren.

Within this marginal zone of the intrusion, even away from prominently xenolithic areas, lithology, texture and grain size vary widely over short distances. Areas of hybridised rocks (for example (S83489)) suggest mixing of dioritic and more granitic magmas.

Moulzie Burn Diorite (HDM)

The Moulzie Burn Diorite [NO 29 79] lies immediately north of the Glen Doll Diorite. On the 1:63 360 scale geological Sheet 65 Balmoral, it is indicated as being part of the Juanjorge intrusion. However, not only are the two intrusions very different in field relations and mineralogy, but magnetic data allow the location of their boundaries to be clearly identified. There is, in fact, a screen of country rock between the intrusions, though this can be shown only by detailed study of the magnetic profiles as there is no exposure in the area.

The Moulzie Burn Diorite is a small but relatively well exposed unit of irregular shape, with many pods and apophyses running into the surrounding Dalradian gneisses. In appearance and mineralogy it is indistinguishable from the diorite portion of the Glen Doll Diorite; it is xenolithic and locally strongly faulted. The two diorite bodies are considered to be part of the same magmatic pulse, if not united at depth. The close spatial association of the Glen Doll Fault with the two diorites suggests that as a line of weakness it acted as a conduit for the ascending magma.

Juanjorge Quartz Diorite (qHDJ)

The Juanjorge Quartz Diorite crops out east of Bachnagaim [NO 2551 7955] and is well exposed in the River South Esk and on the massive crags of Juanjorge and cliffs of Moulnie Craig. It has a gently arcuate lens-shaped outcrop, 5.5 km long and up to 1 km wide in the neighbourhood of Broom Hill [NO 269 798]. In broad terms the Juanjorge intrusion is bounded by Lochnagar L1 Granite to the north and Dalradian metasedimentay rocks to the south. However, at the western end a broad tongue of the L1 granite extends south-eastwards forming the southern boundary of the Juanjorge body for 2 km, although a detached lens of quartz-diorite, 30 to 150 m wide, occurs to the south of the granite in West Corrie and the lower part of East Corrie. At its north-east extremity the body terminates against the Glen Doll Fault.

The Juanjorge body was labelled as coarse-grained granodiorite (G1) by Oldershaw (1974), but recent studies indicate that it lies principally in the quartz–dioritetonalite field (Table 4). The body is of very uniform appearance and lacks the petrographical variety shown by most of the other diorites. It consists almost entirely of a black and white, coarse-grained, biotite-plagioclasedominated rock with a high magnetic susceptibility; values generally lie between 18 and 28 x 10−3 SI. The rock (for example (S83907)) comprises a coarse (2 to 5 mm) aggregate of plagioclase, quartz, biotite, green hornblende and ilmenite/magnetite. The plagioclase ranges from An42 to An33 and is commonly zoned. Biotite is a dark green-brown and contains inclusions of monazite and apatite. The hornblende is generally euhedral but in parts has corroded margins. Sphene is common in this rock as a minor alteration product of ilmenite; elsewhere (for example (S92865)) it is present as subhedral grains up to 1.5 mm across. The rock is widely jointed and altered to a reddish colour only along some joints and small faults, such as in the Burn of Altduthrie, by Bachnagairn. Oldershaw (1974) noted that the Juanjorge body carried an igneous foliation, but this is quite rare; for the most part the rock appears to be unfoliated.

Some variation in the rock is apparent on the east side of Broom Hill with patches of more medium-grained diorite including examples with mafic spots similar to the 'mafic spot' rock of the Abergeldie Complex ((Plate 16)a). In the same area [NO 281 795] there is also a small (120 X 100 m) patch of spheroidally weathered hypersthene gabbro at the contact between the quartz-diorite and the Gourock Granite. The gabbro is veined by granite and is assumed to be an early phase of the Juanjorge intrusion. The gabbro (S92984) consists largely of stumpy to lath-like plagioclase with composition An64 intergrown, in places ophitically, with brown and green hornblende, pink hypersthene and biotite. A few relict clinopyroxene cores indicate that this was the precursor for much of the hornblende. The bulk of the biotite is foxy red and appears to have grown at the expense of hornblende, although some of the larger plates could be primary. An olive to pale green variety of biotite is also present, apparently developing from hypersthene. In places with plagioclase it forms a myrmekite-like symplectite which has grown into the plagioclase.

A fine-grained facies of the quartz-diorite occurs adjacent to the margin of the detached portion of Juanjorge [NO 2594 7850] in West Corrie. In addition to the usual quartz-diorite mineral assemblage, the rock (S82940) includes tremolite which has developed from clinopyroxene. This marginal phase may thus represent an earlier more dioritic phase, equivalent to that seen in xenoliths in the coarse-grained quartz-diorite and more extensively in the L1 granite in the River South Esk and Loupshiel Burn sections.

The contact of the Juanjorge Quartz Diorite with the metasedimentary rocks and amphibolite can be traced along the eastern crags of Moulnie Craig [NO 27 79] and thence over the southern face of Broom Hill. In general it is steeply dipping, but in detail it is cuspate, even convoluted, and large rafts of amphibolite and hornfelsed metasedimentary rock are included within the marginal parts of the body. Limited diorite veining occurs in the metasedimentary envelope. Large xenoliths and screens of quartzose and micaceous psammites with calcsilicate-rocks are also well exposed in the River South Esk section around [NO 2626 7920].

Cul nan Gad and Allt Darrarie quartz-diorites

These two intrusions lie on the outer eastern and southeastern fringe of the Lochnagar L1 Granite. The Cul nan Gad, the larger of the two, forms an arcuate body, 4 to 5 km long and up to 1.8 km wide, which extends northwards from the valley between Creag na Slabhraidh and Hunt Hill, across Glen Muick to the north-west side of the col joining Meall Gorm and Craig Megen. A strong magnetic anomaly points to the drift-covered valley to the north also being underlain by diorite which may be joined at depth to the Cul nan Gad body. The Allt Darrarie body [NO 31 83] is considerably smaller (600 X 400 m) but is better exposed and illustrates well the field relations of these two units.

Both bodies consist largely of medium-grained, inequigranular quartz diorite which in places is foliated. However, on the eastern slopes of Meall Gorm between [NO 305 891] and [NO 316 880] the Cul nan Gad intrusion also includes a screen of granodiorite up to 200 m wide and about 1500 m long next to the Lochnagar Granite. The quartz diorite typically consists of equant to late-shaped plagioclase with intergranular clusters and isolated grains of pale green to brown hornblende, which locally contains relict clinopyroxene. In most of the specimens collected from the Meall Gorm–Craig Megen col the hornblende is partly overgrown by foxy red biotite, which in one sample (S82166) forms 3 to 4 mm long poikilitic plates. Quartz, fine-grained plagioclase and minor K-feldspar occur as intergranular to interstitial phases. Because of its greater K-feldspar content thin section (S82164) is modally a quartz-monzodiorite (Table 4), and contains minor graphic-type intergrowths.

The Cul nan Gad and Allt Darrarie quartz-diorites are both highly xenolithic, though the former is large enough for a non-xenolithic core to be present. On the hill of Cul nan Gad itself [NO 32 87] and on the Meall Gorm–Craig Megen col [NO 31 88] there are many very large country rock xenoliths, and it is likely that they indicate proximity to the top of the intrusion. The Allt Darrarie quartz-diorite is, in places, little more than a vein complex, also representing the roof of the diorite body.

The western margin of both bodies is defined for the most part by Lochnagar L1 Granite, of which numerous veins intrude the diorite. The quartz-diorite veins intrude the country rock extensively along the outer margins of both bodies, so there are broad contact zones. In the north, the Cul nan Gad body is bounded to the east by the same north–south fault that cuts the Glen Mark metadiorite and, though the diorite shows no shearing or amphibolitisation (compare with Glen Mark metadiorite, Chapter 9) it may be that the fault line has acted as a conduit for intrusion.

Abergeldie Diorite Complex

Covering over 30 km2, the Abergeldie Diorite Complex is the largest of the dioritic bodies in the Ballater district. It forms part of a south-east-trending zone of dioritic intrusions that extend from the eastern edge of the Cairngorm Granite to the Crathie area. It is distinguished from the other diorites in the zone by the additional presence of granodiorite and granite. The complex dominates the ground between the north-eastern edge of the Lochnagar Granite and the southern edge of the Glen Gairn Granite both of which are slightly younger. It includes the diorites and granodiorites of the Creag nan Gall area [NO 28 89] to [NO 26 93] which formed part of Oldershaw's (1958) North-east complex of diorite and granite' (the granitic parent is now assigned to the Lochnagar L1 Granite). In marked contrast to most of the other intrusions in the district, the Abergeldie Diorite complex is characterised by a large volume of included Dalradian rock. The enclosed material ranges from centimetre-scale xenoliths to screens and rafts which may be up to 2 km long and 1 km wide. The distribution and disposition of the larger inclusions suggest that most are relatively undisturbed and hence comprise the original roof of the intrusion. The roof zone is well exposed on the cliffs forming the south face of Creag a' Chlamhain [NO 267 954], particularly at the western end where the gentle easterly sloping contact is defined by bedding-parallel joints in graphitic schist. Although in many parts of the complex the topographical highs are occupied by country rocks, the reverse is true in the area around Mains of Monaltrie [NO 242 952], thus emphasising the rolling nature of the roof.

With eleven component rock types, the Abergeldie Diorite Complex shows significantly greater lithological variation and compositional range than any of the other major intrusive bodies. The complex consists principally of moderately coarse quartz-diorite and granodiorite with smaller volumes of medium-grained diorite and quartz-diorite, coarse-grained diorite, tonalite, quartz-monzodiorite, clinopyroxene granodiorite, leucogranodiorite, porphyritic pink granodiorite and granite. In broad terms, there is a preponderance of granodiorite on the eastern side of the complex from The Maim [NO 27 97] to Tom Bad a' Mhonaidh [NO 29 91] and again on the western side between Creag an Lurchain [NO 26 93] and the headwaters of the Girnock Burn [NO 28 89]. Granite occurs only patchily and is confined to the north side of the Dee.

In the field, the relationship between the various phases is not always apparent, particularly since none show evidence of marginal chilling. It is accepted that local variations, such as the clinopyroxene granodiorite, probably result from the assimilation of country rock. However, on the evidence of intrusive relationships at a few critical localities it appears that there is a with-time increase in the acidity of the magma. For example in places medium-grained diorite is enclosed by coarse quartz-diorite [NO 2685 9317] and veined by granodiorite [NO 2455 9761] ((Plate 16)a). Elsewhere, coarse diorite is also demonstrably older than quartz-diorite [NO 2701 9294] ((Plate 16)b) but its relationship with medium-grained diorite is not known. Most enigmatic is the quartzdiorite–tonalite–granodiorite interrelationship as, in places, there appear to be gradational contacts. The only definite evidence is to be found north of Sron Dubh [NO 2926 9693] where medium- to coarse-grained quartzbiotite diorite is cut by granodiorite. The white granite phase of the complex can be seen cutting tonalite to the north of Sren Dubh in grid square [NO 29 96]. In places, close to the margins of the Lochnagar and Glen Gairn granites the dioritic and granodioritic rocks are intensely veined by the earliest phases of the granite complexes.

Field occurrence and petrography

The following section summarises the field and petrographical characteristics of all the main lithological components of the complex. Detailed petrological descriptions of the rocks are contained in Phillips (1991, 1992). Most of the individual areas of outcrop shown on the 1:50 000 and 1:10 000 scale maps are, in fact, characterised by variability, with one or more other lithologies occurring with the mapped unit.

Medium-grained dioritic rocks

These rocks are considered to be amongst the oldest members of the complex. They were recorded at seven localities: the north-east face and summit of Tom a' Chuir [NO 26 93]; the northeast edge of the Creag nan Gall summit plateau due north of the summit [NO 26 91]; immediately south-east of Princess Alice's Cairn [NO 2485 9373]; 150 m south-west of Sron Dubh summit [NO 2907 9631]; Creag a' Chlamhain north-west and east of the triple fence junction at [NO 2702 9587], and less commonly on the southern and western slopes; on the slopes 300 to 600 m north of Rimarsin [NO 263 965] and east of the Crathie Burn [NO 246 976]. With the exception of the occurrences north and east of Creag a' Chlamhain where the medium-grained rock predominates, these rocks tend to occur as screens within areas dominated by coarse quartz-diorite or granodiorite. On Sron Dubh the enveloping rock is white granite (the final phase of the complex), whereas at Princess Alice's Cairn the dioritic rock occurs as a 2 m-wide screen within L1 Lochnagar Granite, and can be traced for almost 100 m. In most places contact with the surrounding rocks is largely obscured, but on Tom a' Chuir [NO 2694 9313] the medium-grained diorite is veined by coarse quartz-diorite which in turn is cut by granodiorite. The medium-grained dioritic rocks comprise two megascopically distinctive types.

Porphyritic quartz-diorite

Most evident, though not necessarily most widespread, this lithology comprises a pale to mid grey, medium- to fine-grained rock studded with rounded to angular, dark green spots which earned it the field sobriquet 'mafic spot' rock ((Plate 17)a). The spots may be up to 4 mm in diameter and comprise clots of fine-grained anhedral hornblende partly mantled or enclosed by anhedral brown biotite (S92666). Biotite and finely divided opaque oxide also occur within the clots. The shape of the clots, which in places have noticeable euhedral outline ((Plate 17)a), suggest that some may be after glomeroporphyritic aggregates whereas others are simply replacing single crystals. Somewhat smaller (1 to 2 mm) plagioclase phenocrysts which enclose small rounded biotites are present in some samples. The remainder of the rock consists largely of anhedral to weakly subhedral plagioclase intergrown with subordinate stumpy biotite and hornblende. Plagioclase, quartz and K-feldspar all occur as anhedral intergranular phases. Sphene, apatite, opaque oxide and zircon are present as accessories.

Equigranular diorite, quartz-diorite and tonalite

These rocks (for example (S92665)) have an essentially similar mineralogy to the 'mafic spot' rock (Table 4) but are distinctly more equigranular and only weakly porphyritic. They comprise an aggregate of equant and tabular plagioclase with intergranular biotite and hornblende, with some plagioclase and, more rarely, green-brown hornblende phenocrysts up to 3 mm long. Plagioclase, quartz and, less commonly K-feldspar occur as late crystallising intergranular phases. The volume of quartz and K-feldspar is more variable than in the 'mafic spot' rocks such that the suite includes quartz-diorites and tonalites and even encompasses rocks which approach granodiorite. In most rocks examined, hornblende is the dominant mafic phase, although in a few instances this role is held by biotite, most notably in thin section (S92663) where it occurs to the total exclusion of hornblende. Notable variants in terms of mafic minerals are thin sections (S92916) and (S94295), where foxy red-brown biotite predominates, clinopyroxene is commonly present and pale green hornblende is only rarely seen.

The relationship between the porphyritic and equigranular varieties has not been established. On the northern side of Tom a' Chuir [NO 2694 9313] [NO 2685 9317] both occur in close proximity within the same body of coarse-grained quartz diorite. Approximately 100 m to the east, however, the equigranular type appears to grade into the enveloping coarse-grained rock.

The equigranular diorite, which forms a 15 m-wide dyke-like body on the 470 m summit [NO 2685 9613], includes angular centimetre-scale xenoliths of fine-grained and weakly porphyritic hornblende-biotite microdiorite. Excepting the reduced grain size, the absence of quartz and a higher proportion of hornblende, the xenoliths (for example (S82352)) are petrogiaphically similar to the host with coarse sphene being a feature of both rock types. The microdiorite would therefore seem to represent an early phase of the Abergeldie Diorite Complex and not an inclusion of the adjacent late tectonic basic intrusive.

Coarse-grained dioritic rocks

Diorite

In terms of areal extent this is the smallest component of the Abergeldie Diorite Complex. The most extensive outcrop (about 500 X 220 m) occurs between the upper reaches of the Coulachan and Crathie burns, underlying much of a low north–south-trending ridge centred on [NO 246 981]. Because of its high colour index (locally greater than 50), the rock is readily distinguishable from the surrounding and intruding quartz diorite and granodiorite, but has not been mapped separately at 1:50 000 scale. It comprises (S92921) a coarse to very coarse, intergrowth of (in part poikilitic) hornblende and andesine in the range An38–44. Hornblende occurs both as large single crystals and coarse aggregates; in places these have been replaced by aggregates of smaller irregular crystals. A noticeable feature of hornblende in thin section is the colour zonation within individual crystals from brownish green through olive green to sea green. Inclusions of brown biotite and rods and granular trails of opaque oxides are commonly encountered. Quartz is a minor interstitial phase. Apatite, as particularly coarse inclusions within plagioclase and quartz ((Plate 17)b), coarse euhedral sphene, opaque oxide and zircon are accessories. Thin section (S92919) is a more homogeneous rock with a similar mineralogy, but with the additional presence of probable cummingtonite.

Diorite is present elsewhere within the complex, but since the colour index is considerably less than 50 (Table 4), the rocks are not easily distinguished from the more widespread quartz-diorite. On the evidence of thin sections, diorite also occurs on the southern slopes of Creag a' Chlamhain at [NO 2674 9536] and [NO 2655 9526], on Tom Buailteach [NO 2766 9360] and on Tom a' Chuir [NO 2701 9274]. Two Creag a' Chlamhain samples ((S92670) and (S92976)) are medium- to coarse-grained, inequigranular, holocrystalline rocks with a pronounced foliation. The fabric is defined by length-orientated, zoned and twinned plagioclase laths, large (up to 2.5 mm) biotite flakes and chains of smaller brown-green hornblendes. Sphene forms anhedral to euhedral crystals up to 1.0 mm in length, which are commonly aligned along the foliation. Minor quartz and K-feldspar occur as intergranular to interstitial phases. The only significant difference between the two rocks is that in one (S92670) biotite is foxy red-brown as opposed to brown, and apatite needles are distinctly cloudy.

Quartz-diorite and tonalite

Quartz-diorite is one of the main lithologies of the Abergeldie Diorite Complex; with minor tonalite and rare quartz-monzodiorite, it forms the major part of the complex and much of the exposed bedrock. The rock is well exposed in a number of areas, most notably on Tom a' Chuir, Creag a' Ghobhainn to Craig Gowan (immediately south of Balmoral Castle), the southern and western slopes of Creag a' Chlamhain, Craig Nordie and in the upper reaches of the Crathie Burn between Creag Mhór and Coulachan Burn.

For the most part the quartz-diorite is a uniform coarse-grained black and white rock, although medium-grained and weakly porphyritic varieties are also known. Typically it comprises an inequigranular aggregate of plagioclase, hornblende, biotite and quartz with minor K-feldspar and accessory sphene, apatite, zircon and opaque minerals. Plagioclase, which generally makes up between 40 and 50 per cent of the rock, forms suhedral equant to lath-like zoned crystals with cores ranging from An40 to An54 zoned to rims of An31. Hornblende and biotite exhibit a range of forms including small rounded chadacrysts within plagioclase, single crystals which may be intergranular to plagioclase, and large clusters of grains some showing an ophitic-like inter-growth with plagioclase. The two minerals are, in general, closely associated, hornblende being mantled by or enclosing biotite. Hornblende is mostly, but not always, the dominant mafic mineral. For example, in thin section (S93072) aggregates of very pale green hornblende are distinctly subordinate to coarse plates of foxy red-brown biotite, both minerals having apparently grown at the expense of clinopyroxene. In thin section (S94301), a medium to coarse, leucocratic quartz-diorite, foxy red-brown biotite is the only mafic mineral present.

Quartz, plagioclase and K-feldspar occur as late crystallising intergranular phases. More rarely K-feldspar forms larger crystals (up to 3 mm) which poikilitically enclose finer grained plagioclase, hornblende and biotite. Oldershaw (1958) noted that some of the coarse K-feldspar might be metasornatic and derived from the L1 granite, for example that by the road south of Craig Gowan. In common with most other lithologies within the complex, sphene is characteristically coarse (1 to 2 mm), euhedral and is associated with the mafic clusters.

Some of the more quartz-rich rocks of this suite plot within the tonalite field (Table 4), but as the difference between them and the quartz-diorite is quite subtle, a megascopic distinction in the field is not possible. However, on the evidence of thin sections, tonalite is the dominant lithology in the area contained by the path leading south-west from Prince Albert's Cairn [NO 260 935] and the northern end of Creag nan Gall. It is present to a lesser extent within quartz-diorite on Tom a' Chuir, Creag a Ghobhaim [NO 251 940], Craig Nordie, the southeast slopes of Cam Moine an Tighearn [NO 2347 9701], on the south and east sides of Creag a' Chlamhain, Torgalter Burn at [NO 2829 9573] and [NO 2872 9572], within the Glen Gairn Granite about 500 m south-south-east of Cnoc Chalmac [NO 266 006] and within the Lochnagar L1 Granite below Princess Alice's Cairn [NO 249 937].

But for the increased proportion of quartz, the generally slight dominance of biotite over hornblende and the presence of microcline within the K-feldspar component, tonalite is petrographically comparable with quartz-diorite. Quartz and K-feldspar remain as intergranular phases. In thin sections (S82172) and (S92663) biotite is the only mafic mineral present.

Granodiorite

Occupying a combined area of over 20 km2, granodiorite in its various forms is the other main component of the Abergeldie Diorite Complex. The largest outcrop, on the south-west slopes of Geallaig Hill, extends from the area of The Maim [NO 269 968] for over 2 km to south of Abergeldie Castle. A second extensive granodiorite body can be traced for 4 km south-west from the area of Prince Albert's Cairn [NO 260 934] and, with a width of up to 0.7 km comprises the bulk of the Creag nan Gall ridge. The Creag nan Gall Granodiorite is sufficiently distinct from the other granodiorites to be treated separately on the 1:50 000 series map. Somewhat smaller but equally discrete bodies are centred on Tom Bad a' Mhonaidh [NO 287 920], Bush, Crathie [NO 256 964] and the lower north-east slopes of Craig Nordie [NO 235 948]. However, there is also a considerable volume of granodiorite in the form of minor intrusions within diorite and tonalite which are too small to be delineated on the 1:50 000 map.

The constituent rock shows considerable range in texture, grain size and colour, which in part may reflect contamination by included country rock and/or different intrusions. Lensoid screens of metasedimentary rock are particularly evident on Tom Bad a' Mhonaidh where they are largely aligned parallel to a very pronounced and pervasive fabric in the granodiorite (Plate 18), defined principally by orientated feldspar megacrysts. The most commonly encountered rock is coarse to medium grained, inequigranular to porphyritic, with megacrysts of plagioclase and, more rarely, K-feldspar up to 5 mm across. The K-feldspar may well be metasomatic in origin (Oldershaw, 1958). In thin section the rock is seen to comprise an open network of subhedral plagioclase, biotite and hornblende enclosed and infilled by an intergrowth of poikilitic quartz, K-feldspar and plagioclase. In one or two instances the quartz content is sufficiently depressed for the rock to be classified as a quartz-monzodiorite. Clinopyroxene-bearing rocks were recorded at localities north of the River Dee [NO 2479 9459], [NO 2557 9434] and [NO 2419 9767] ((Plate 17)c). Though its shape is largely obscured by overgrowths of biotite, hornblende or uralite, the clinopyroxene appears to have been largely granular or intergrown with plagioclase. The fact that one occurrence lies close to an outcrop of clinopyroxene-rich calcsilicate-rocks suggests that this could be further evidence of contamination. Typically, coarse-grained subhedral sphene, iron oxide, zircon ((Plate 17)d) and apatite are invariably present as accessories.

The Creag nan Gall Granodiorite is more uniform in appearance than the other granodiorite bodies with only one small area of dioritic rock present. The granodiorite is particularly distinctive in the field; it is a comparatively leucocratic white and black rock which is characterised by coarse bladed biotite (up to 8 mm long) and macroscopic sphene ((Plate 17)e). The unusual form of the biotite appears to be a primary feature, there being no evidence that it has pseudomorphed hornblende. Hornblende, where present, is generally subordinate to biotite (Table 4).

An equally distinctive but less commonly encountered variety of granodiorite was recorded near Albert's Cairn [NO 2605 9323], and at the western end of Tom a' Chuir [NO 2672 9255]. This rock is typically inequigranular with white plagioclase phenocrysts, up to 5 mm across, in a medium-grained pink matrix flecked with biotite and rare hornblende. The matrix comprises finer grained plagioclase, K-feldspar and quartz. Late rounded quartz blebs are locally developed, producing a coarse graphic-type intergrowth. Overall the proportion of quartz and K- feldspar seems greater than in other granodiorites. The pink variety occurs within, and has sharp contacts with, the Creag nan Gall type; the fact that both have bladed biotite may indicate some genetic link.

Granite

Granite is not a major constituent of the Abergeldie Diorite Complex, the only outcrops of any size being those forming Sron Dubh [NO 29 96] and The Maim [NO 28 97]. Smaller exposure are present below the northern pier of Balmoral Bridge [NO 2622 9492] and in the small disused quarry on the east side of Craig Nordie [NO 2362 9473]. The Craig Nordie occurrence is in an area dominated by granodiorite, but no contact between the two rock types is seen. The sides and roof of the quarry, and inclusions within the granite are entirely of metasedimentary rocks consisting largely of idocrase-rich calsilicate-rock with subordinate sulphide-enriched schist. The country rock is intruded by granite veins up to 130 mm wide with narrow (10 to 30 mm) biotite-free margins which may represent a chill feature. The erosion level on Sron Dubh is close to the original roof of the intrusion with much of the summit ridge comprising Blair Atholl and Easdale subgroup rocks. These and the adjacent tonalite are sporadically veined by white granite.

The Abergeldie Diorite Complex granite is distinct from other granites in the Ballater district in being white, or even bluish, leucocratic and mostly equigranular. The rock consists (Table 5) essentially of coarse perthitic K-feldspar, slightly smaller plagioclase and pools of interlocking euhedral quartz. The K-feldspar:plagioclase ratio is quite variable, particularly in the Sron Dubh body where monzogranite is not uncommon and some rocks approach granodiorite in modal composition. Foxy red-brown biotite is the only mafic mineral present; accessories include iron oxide, zircon, cloudy apatite ((Plate 17)f and more rarely sphene.

The texture, rock and biotite coloration, low magnetic susceptibility and high inclusion content of the apatite are the features which distinguish this granite from the Lochnagar L1 and Glen Gairn granites, although their overall compositions are generally similar. The age relationship between the three granites cannot be established with absolute certainty since nowhere are they seen in contact. On the evidence that the white granite is, in places, cut by grey microgranite dykes that are never seen to intrude the Glen Gairn Granite, it would seen reasonable to suggest that the white granite is the older. This conclusion is strengthened if the presence of foxy red biotite and cloudy apatite are accepted as being indicators of thermal metamorphism (information from A J Highton, BGS, 1999).

Foliation in dioritic rocks

Oldershaw (1974) described the Lochnagar L1 granite and G1 granodiorite as being foliated, and his (Figure 1) indicates that this fabric is widely developed in these rocks and is essentially arcuate. The conclusion reached during this survey is that a foliation is largely absent from the granite, but is present to a limited extent in some of the granodioritic and dioritic rocks. In terms of specific intrusions the Juanjorge, Glendoll and Moulzie Burn diorites and all the granitic bodies are unfoliated. The Abergeldie Complex rocks show sporadic development of a fabric particularly within granodiorite, most notably on Tom Bad a' Mhonaidh where it is especially pervasive (Plate 18), and on the south-east flank of Tom a' Chuir [NO 26 93] where it occurs more patchily. There is more widespread development of foliation in the Cul nan Gad and Allt Dararrie bodies, but even there it is not ubiquitous. In the latter two bodies the fabric is defined by the orientation of mica laths and a slight shape fabric in the quartz. On Tom Bad a' Mhonaidh, it is most apparent in the common alignment of feldspar megacrysts and, to a lesser extent, in hornblende and biotite. Where inclusions of country rock are present they are, irrespective of size, also aligned in the plane of the foliation, the smaller examples showing indications of flattening and being drawn out into schlieren.

The foliation in the Cul nan Gad and Allt Darrarie bodies is mostly near-vertical and parallels the contact with the Lochnagar Granite such that in the former it has an arcuate trend. The orientation in the Abergeldie rocks is more variable and complex, ranging in strike from north–south to east–west and from vertical to sub-horizontal. However, with the possible exception of certain alignments in the area of Tom Bad a' Mhonaidh, most of the foliation direction relates to the margins of the Lochnagar and Glen Gairn granites.

The overall impression given is that this is not a flow fabric, but a tectonic flattening imposed on a rock which had not completely solidified, possibly induced by the intrusion of the early Lochnagar and Glen Gairn granites. This conclusion is largely based on the observation in the Cul nan Gad Diorite that the fabric intensifies towards the contact with the L1 granite (Goodman et al., 1990). On Tom Bad a' Mhonaidh, a co-parallel foliation is also evident in small bodies of L1 granite which intrude the granodiorite. This may relate to emplacement of the L3 granite bodies present in the area.

Chapter 13 Post-tectonic major intrusions—granites

Lochnagar Granite

The Lochnagar granite pluton crops out over an area of approximately 150 km2, of which about two-thirds lies on Sheet 65E Ballater and the remainder on Sheet 65W Braemar (Figure 20). The ensuing description of the granites is based on the work of Barrow and Cunningham Craig (1912), Oldershaw (1958, 1974) and some modifications to Oldershaw's mapping and interpretation made as a result of reconnaissance mapping during this survey. The main changes are as follows:

The overall picture which emerges is one of a nested set of granite plutons. The earliest phase was a roughly circular pluton, about 15 km across, of the L1 granites which was subsequently cored out by a 12 X 9 km pluton of L2 granites, which breached the L1 granites in the north-west. Both L1 and L2 granites are intruded by microgranites of similar composition to their hosts. The L3 granites are the most evolved chemically and occur at a number of places intruding the L1 granites; their relations with the L2 granites are unknown but none is at present known from within the L2 outcrop. Unlike the red granites of Ballater and Mount Battock, those of Lochnagar are grey and pink or pink and exhibit a range of magnetic susceptibilities. Consequently, the aeromagnetic anomaly pattern of the Lochnagar area closely reflects the surface distribution of the various granite (and associated diorite) phases (Figure 3).

L1 granites (GL1)

The L1 granites form a crescentic outcrop around most of the granite body extending into the Braemar district (Sheet 65W). For the most part the rocks are coarse and porphyritic, but in the south-east quadrant they are coarse and inequigranular. A sharp, non-chilled contact between these two variants is exposed on a loose block east of Creag nan Gall [NO 2790 9192]. There is also some variation among the porphyritic varieties which may indicate that they are the product of separate intrusive pulses.

The rocks are characteristically coarse grained (2 to 4 mm). Oligoclase (forming about 35 per cent) is the dominant constituent (Table 5), occurring as white equant to lath-shaped subhedra; these are fairly fresh but may have the cores of zoned crystals patchily altered to sericitic mica. K-feldspar (about 30 per cent) is mostly untwinned orthoclase perthite; it occurs as pink anhedra, ranging up to about 10 mm long in the inequigranular varieties and up to about 15 mm long with subhedral form in the porphyritic varieties. There are some small developments of myrmekite. The larger crystals are commonly poikilitic, enclosing quartz, plagioclase and biotite. The quartz (about 25 per cent) appears greyish and occurs as intergranular to interstitial anhedra or in small clusters. Biotite (6 to 9 per cent) is typically dark khaki but greenish in some thin sections, and appears to be seriate between tiny subhedral specks and euhedra up to 4 mm across; it is mostly fresh but in places is partly altered to chlorite. Green hornblende is generally present in only trace amounts and as skeletal remnants. It may also occur as subhedral to acicular crystals (up to 4 mm long) that form up to about 1 per cent of the rock and are partly replaced by biotite. Zircon, apatite and sphene occur both as discrete grains and as inclusions in biotite; sphene, as euhedra up to 2 mm across, is the most common and largest of these.

Small mafic xenoliths in various stages of digestion are sparsely distributed throughout the L1 granites, but are common in places near the southern and south-eastern margins. Oldershaw (1958) recorded that they are mostly discoidal in shape and generally orientated parallel to the margin of the granite; this he attributed to flowage in the granite. The granite itself, however, is generally unfohated, but in places there is a weak fabric, most commonly shown by some preferred orientation of the larger feldspars. The foliation and orientation of xenoliths is probably attributable to a slight ballooning of the granite, the pressure causing some steepening of foliations in the country rocks (Oldershaw, 1958), and inducing a flattening foliation in the marginal Cul nan Gad and Allt Darrarie diorites (Goodman et al., 1990).

The L1 granites clearly postdate the various fringing diorites and granodiorites, but the nature of the contact varies from place to place. In the area south-east of Balmoral, Li is unchilled against Abergeldie Complex granodiorite; the contact is generally sharp, very irregular and of variable dip, and a few veins of Li granite extend outwards from the main mass. The L1 granites clearly postdate the more basic rocks, and what at first appear to be xenoliths of L1 coarse granite in granodiorite and tonalite east of Creag nan Gall [NO 2798 9300] and [NO 2790 9192] are probably remnants of granite veins.

The Juanjorge Quartz Diorite almost certainly predates most of the Lochnagar granites, although its relationship with the L1 granite is not always clear cut. In the River South Esk and nearby, Barrow and Cunningham Craig (1912) described a 'very perfect passage' between L1 granite and a tonalite facies of the Juanjorge Quartz Diorite over a distance of about 300 m. However, localised chilling of the granite (Oldershaw, 1958) and thermal metamorphism of the diorite (Jarvis, 1987) imply a time lapse between the two intrusions. The eastern part of the northern Juanjorge/L1 contact has been reinterpreted by Goodman et al. (1990) as a zone of contamination and hybridisation consequent upon the L1 granite having been intruded when the tonalite was still partly mobile. In the detached outcrop of the Juanjorge Quartz Diorite near the Burn of Loupshiel [NO 2645 7881] coarse quartz-diorite is cut by near-vertical, east–west-trending veins of coarse granite which show considerable signs of hybridisation near the contact. Nearby in West Corrie [NO 2544 7878] the diorite is cut by more or less vertical sheeted veins of granite which link with the main Li outcrop to east and west. A 25 m-thick lens of quartz-diorite is included in a thick sheet of granite exposed in the headwall of East Corrie [NO 2611 7821], and several metres of 'hybrid' granodiorite have developed at the contact.

In contrast, a sharp contact between L1 granite and Juanjorge Quartz Diorite is exposed in the Burn of Altduthvie [NO 2536 7954]. The granite shows localised reduction in grain size adjacent to the contact. Likewise, at the contact against the Allt Darrarie Diorite there is some chilling of L1 and inclusion in it of sharp-edged diorite blocks.

L2 granites (GL2 )

These rocks form an oval pluton extending for 12 km north-westwards from Loch Muick to just across the River Dee. The southern part of the pluton is moderately well exposed, although there are extensive blockfields on many of the rather rounded higher peaks, and spreads of glacial deposits and peat lower down. Much of the central and eastern parts of the pluton, particularly in Glen Gelder, is obscured by a variety of glacial and outwash deposits.

The L2 granites are mostly medium to coarse grained, inequigranular, unfoliated grey and pink leucocratic rocks. Characteristically, quartz occurs partly in small clusters, and biotite occurs as scattered isolated flakes, commonly of very small size. No subdivision of the L2 granites has been mapped, although Oldershaw (1958, 1974) recorded a number of variants which probably represent separate intrusive pulses. While most of the rocks are inequigranular, they grade by a modest increase in the size of the K-feldspars into a porphyritic facies; Oldershaw recorded the latter around Cac Carn Beag [NO 242 863] and Meikle Pap [NO 260 860], and it also occurs south of Creag nan Gall [NO 267 907].

The major constituents are quartz (28 to 35 per cent) and orthoclase-perthite (25 to 38 per cent). The former occurs as equant subhedra and anhedra up to 4 mm across and as clusters up to 8 mm across. The perthite forms pale pink anhedra mostly up to 5 mm, but up to 15 mm long in the porphyritic facies; the phenocrysts are commonly mildly poikilitic. Oligoclase (20 to 40 per cent) is white and clouded from partial alteration to sericite, and is of subhedral to anhedral form. There are some rounded phenocrysts of plagioclase up to about 15 mm. A little myrmekite is present in some samples. Within a single sample biotite (4 to 6 per cent) ranges from subhedra, up to about 4 mm across, down to scattered tiny flakes, and from completely fresh to largely altered. The fresh biotite is a khaki brown colour in thin section, but it commonly shows minor alteration to chlorite, while more altered crystals are replaced by chlorite and sphene ± epidote. Accessory constituents are opaque oxide that form small euhedra, and subhedra with small clusters in parts, sphene as separate euhedra, some allanite euhedra, and small zircons and apatites occurring as both separate euhedra and as inclusions in biotite.

One textural variant occurring south-west of Princess Royal's Cairn [NO 236 924] is a more uniformly medium-coarse rock with scattered K-feldspar phenocrysts up to 15 mm long. Another variant formerly quarried [NO 2407 9266] to the south-east of the Princess Royal's Cairn, contains sparse but conspicuous white plagioclase phenocrysts and small pink orthoclase-perthite phenocrysts. The groundmass is medium to fine-grained having subhedral quartz and biotite that is subhedral or bladed in habit. The rock is two phase, with equigranular micro-granite trailing between many of the larger crystals. This microgranite appears to have been in place at an early stage since components of it are enclosed in the outer zones of both the plagioclase and K-feldspar.

L3 granites (GL3 )

The L3 granites are mostly pink equigranular leucocratic biotite granites (although the largest body is white, and occurs in the Braemar district); they are chemically the most evolved rocks on Lochnagar. They clearly intrude the L1 granites but their relations with the L2 granites are not known; none is known from within the L2 outcrop, which might suggest that they predate L2, although radiometric age determinations indicate otherwise (see below).

The L3 granites are known from five widely separated locations in the L1 granites (Figure 20), each differing somewhat in grain size, magnetic susceptibility and, to some extent, colour. Those in the Ballater district on Carn an Daimh [NO 297 871] and at the south-east end of Loch Muick [NO 270 815] are similar pink, medium to coarse, equigranular rocks with bladed biotites, originally linked by Oldershaw (1958, 1974), and interpreted as a ring dyke. These two areas of outcrop are separated by 5 km occupied by a drift-covered slope and by Loch Muick, so the linkage is unproven; the irregular nature of most of the boundaries of the L3 bodies suggest that such a connection is unlikely. The contacts on the main body on Can an Daimh are steeply dipping; the inner contact dips steeply outwards, towards the east. However, there are also a number of sheets of L3 granite to the west of the main body; these cut the L1 granite and have various orientations. The contacts of these L3 granites are not chilled, but both the L3 and adjacent L1 granites are cut by very fine, pink, quartz-feldspar-(biotite-) phyric microgranite which may be a chilled late phase of the L3 intrusion.

The L3 granite on Crean nan Gall is petrographically similar to that at Carn an Daimh, but a little finer grained and it occurs in a number of small and irregularly shaped outcrops.

Microgranite (FL)

The coarse granites of Lochnagar are cut by a small number of fairly substantial bodies of microgranite, most of which appear to be compositionally related to their hosts, and so may belong to the same intrusive episode as their hosts. Aplitic and pegmatitic varieties occur but are generally rare, and quartz veins very rare, indicating that the magmas were rather dry.

Oldershaw (1958, 1974) mapped a substantial body of microgranite cutting the L2 granites around the peak of Lochnagar [NO 250 850] and cropping out over about 3.5 km2. It was emplaced in two pulses (his A4 and A5 adamellites), the later one having a 30 to 60 m chill zone. Very fine microgranite, similar to that in the chill zone, penetrates the first phase of microgranite as horizontal veins 1.5 to 6 m thick. The rock is a pink quartz-orthoclase perthite-plagioclase-phyric microgranite compositionally similar to the L2 granites. In Corrie Nan Eun [NO 230 871] Oldershaw (1958) recorded that the contact of the microgranite with the L2 granites is flat-lying and marked by 15 to 30 cm-thick zones of white quartz-pink feldspar pegmatite which can be traced for about 400 m.

Barrow and Cunningham Craig (1912, p. 84) recorded more or less vertical veins of microgranite and aplite as being 'frequent and are to be seen in nearly every locality where a good section of granite is exposed; in the corries on the northern face of Lochnagar and on the crags to the north and south of the Dubh Loch these veins are often very conspicuous'. Such veins appear, however, to be uncommon around the northern and eastern exposures of the granites. They also recorded that such veins commonly carry muscovite rather than biotite and have central drusy pegmatite cavities in which the quartz is brown or yellow cairngormite. Oldershaw (1958) recorded similar drusy cavities in the microgranites on the summit of Lochnagar, near to their vertical side margins. Barrow and Cunningham Craig (1912) noted that the great corries of Lochnagar and the flanks of Loch Muick displayed sharply cut and precipitous gullies following either microgranite veins or small faults and crush lines which may contain small reefs of quartz. The better-marked quartz reefs run in a north-easterly direction. Overall, however, there are very few quartz veins in the Lochnagar granites.

In contrast there are relatively few microgranite dykes cutting the granites. The only minor intrusions cutting L1 of the Ballater district that are known are in the north-east, where [NO 2595 9363] there are pink and grey quartz-feldsparphyric microgranites; they may be related to similar material in north-east-trending dykes cutting the L3 granites nearby [NO 2583 9377]. A small north-west-trending microdiorite dyke cuts the L1 granite [NO 2560 9438]. Small pink veinlets of quartz-feldspar-biotite porphyry cut the L3 granite and adjacent L1 granite on Cam an Daimh [NO 295 872]; they are probably related to the L3 intrusion there.

Form of the intrusions

Geophysical modelling of the magnetic data suggests that the contacts of both granites and the marginal diorites are more or less vertical. This has to be coupled with field observations which indicate that parts of many of the intrusions have flat-lying roofs within or only slightly above the present topography. There is visible evidence in the Ballater district of rocks roofing the Glen Doll, Allt Darrarie and Abergeldie diorites, the L1 granite at many places in the north-east part of the Lochnagar pluton, and the microgranites near Lochnagar summit. Proximity to the roof of the Cul nan Gad Diorite is indicated by the metasedimentary rafts on Cul nan Gad [NO 32 87]. In addition there are indications of essentially flat-lying contacts of the L1, L2 and L3 granites with Dalradian rocks at various places in the adjacent Braemar district (Sheet 65W). The only major body for which there is no positive indication of proximity to the roof is the Juanjorge Diorite.

Gourock Granite

This small (350 X 300 m) granite body occurs on the southern edge of the Juanjorge Quartz Diorite between Broom Hill and Dog Hillock. Centred on Gourock lochan [NO 282 795], the granite is well exposed on the crags to the east of the lochan. On the western side of the valley, tongues of granite cut hypersthene gabbro of the Juanjorge body and extend into the quartz-diorite for distances in excess of 100 m.

For the most part the Gourock Granite comprises a pink, coarse-grained, equigranular, biotite granite, although locally it may be porphyritic. The rock typically comprises plagioclase (locally dominant), K-feldspar and quartz. Biotite is present as localised large crystals, singly or in aggregates of two or three, which are heavily chloritised and may be associated with coarse magnetite. Thin section (S92868) contains a single, somewhat skeletal grain of pale green amphibole. Apatite and sparse coarse sphene are also present.

The Gourock Granite is separated from the main outcrop of the Lochnagar L1 granite by over 800 m of diorite and by a comparable thickness of diorite and metasedimentary rock from the tongue of L1 on the south side of Juanjorge. Excepting the presence of amphibole, the colour, petrography and the geochemistry (Chapter 14) suggest a closer association with Lochnagar L3 granite than L1.

Khantore Granite

The Khantore Granite is a roughly circular body occupying an area of just over 2 km2 on the south side of the Dee valley between the Geldie Burn and the Creag nam Ban ridge. The granite is moderately well exposed in the Den Burn and Genechal Burn sections, around Khantore [NO 288 938] and Genechal [NO 288 930] and on the slopes of the ridge joining Creag nam Ban and Sgor an h-Iolaire. Other than in the west, the margin of the granite can be fixed with some certainty, although the actual contact is seldom exposed. The southern boundary is against Lochnagar L1 granite and Abergeldie Complex diorite; elsewhere the host is exclusively Dalradian metasedimentary rocks. The eastern contact appears relatively flat-lying, coinciding roughly with the 450 m contour over much of its length. The metasedimentry rocks on the upper part of the ridge form a roof to the granite and as such are strongly hornfelsed and cut by numerous bosses and irregular dykes of granite, one of which can be followed for nearly a kilometre. The frequency and size of these minor intrusions decrease eastwards, the farthest being recorded over a kilometre from the eastern contact. The impression gained is of a contact sloping gently to the east.

Compared with other granites in the area Khantore is relatively uniform in appearance. It consists largely of a medium- to coarse-grained, equigranular, pink or white granite. The colour would seem to reflect the degree of alteration of the feldspars. Secondary reddening is also evident on some joints; there are also localised areas in which the granite is enriched in goethite and limonite as the result of weathering. The main petrographical variation seen in the field is in the proportion of biotite. Some compositional variation occurs in the apophyses in the roof zone, most notably south-east of Creag nam Ban around [NO 3007 9414] where typical Khantore granite, slightly porphyritic biotite-enriched pink and white granite and porphyritic leucogranodiorite occur in close proximity.

The typical Khantore granite consists of quartz and coarsely perthitic K-feldspar, including minor perthitic microcline, intergrown with and enclosing plagioclase. Sub-solidus plagioclase rims and overgrowths and myrmekite are present along K-feldspar grain boundaries. Both feldspars are dusty brown in colour due to limited sericitisadon. Scattered through the rock are subhedral to distinctly ragged anhedral biotites which are variably chloritised and range in colour from deep brown to reddish brown or, more rarely (S92901) green to orange brown. Accessories include zircon, apatite, magnetite, hematite and sphene.

Compositionally and mineralogically the Khantore Granite is similar to the Lochnagar L3 granite, although it lacks the bladed habit common to the biotites of the latter.

Glen Gairn Granites

The 'Glen Gairn Granite Complex' was described by Harrison (1987) in terms of four components: diorite, phase I and phase II granites and zinnwaldite microgranite. In this account the diorite and small portions of the phase I granite around Sron Dubh [NO 29 96] and The Maim [NO 27 97] are assigned to the Abergeldie Complex. In addition, it was recognised by Webb et al. (1992) that the south-eastern part of the phase II granite also contains lithium micas so together with the zinnwaldite microgranite of Gairnshiel it has been separated off as the Coilacriech Granite. The Glen Gairn granites of this memoir now comprises Harrison's phase I granite and the remainder of his phase II granite, referred to here as the Glen Gairn Granite and the Glen Gairn Leucogranite respectively. These two granites occupy around 10 km2 in the Ballater district (Sheet 65E) and are part of a much larger mass which extends for a farther 3 km to the west (Braemar district, Sheet 65W) and 5 to 6 km to the north (Glenbuchat district, Sheet 75E).

Glen Gairn Granite (GG)

The Glen Gairn Granite dominates the north-west corner of the Ballater district. Small areas are quite well exposed such as east of the Crathie–Gairnshiel road around [NO 265 980], south and west of Blairglass [NO 35 99], around Duchrie Burn [NO 23 99] and south of Dalnabo around [NJ 304 005]. However, large areas are poorly exposed, with extensive drift cover in the valleys. The southern boundary is largely formed by Abergeldie Complex diorite and granodiorite, whereas to the east the granite passes below the Dalradian metasedimentary rocks which form the upper slopes of Geallaig Hill [NO 298 982]. The present erosion level appears to lie close to the original roof of the intrusion. Thus the eastern part of the granite, to the north of Geallaig Hill, comprises a sheet and vein complex in the roofing metasedimentary rocks, whereas farther west larger areas of granite are interspersed with areas of sheet and vein complex. The granite is highly xenolithic with inclusions of both diorite and metasedimentary rocks. Anastomosing granite sheets preserve xenoliths in various stages of separation from the host rock; many small xenoliths are reorientated whereas larger inclusions of metasedimentary rock apparently preserve their pre-granite orientation and may represent roof pendants. Some xenoliths show evidence of partial assimilation.

The Glen Gairn Granite is typically medium to coarse grained with a white aspect to the quartzofeldspathic component of the rock. Parts of the intrusion are megacrystic with pink or white orthoclase megacrysts locally over 20 mm long. The granite is composed of plagioclase, orthoclase and quartz in approximately equal proportions together with about 10 per cent of mafic minerals. The latter comprise extensively chloritised yellow to brown biotite, with or without green hornblende. Sphene and apatite are abundant accessory minerals along with opaque minerals, zircon and allanite. As indicated above, Harrison (1987) regarded the Abergeldie-type white granites of Sron Dubh and The Maim as part of his phase I body, but they are readily distinguished on the basis of their equigranular texture, lower colour index, lack of hornblende, foxy red-brown biotite and cloudy apatite.

Glen Gairn Leucogranite (iGG)

The outcrop of the Glen Gairn Leucogranite is largely within the boundaries of the Glenbuchat district (Sheet 75E) but a small tongue extends south to the B976 road north-east of Braenaloin [NJ 28 00]. According to Harrison (1987), phase II comprises a coarse, pink, biotite granite very similar in appearance to the Cairngorm or Monadhliath granites. Medium- to coarse-grained, slightly porphyritic pink granite, which texturally resembles the Lochnagar L3 granite and may well relate to the leucogranite, occurs patchily within the Glen Gairn Granite close to its southern edge south of Blairglass [NO 258 994]. The rock there (for example (S77680)) (Table 5) is dominated by coarsely perthitic K-feldspar which includes minor microcline, and poikilitically encloses plagioclase and rounded quartz. In this specimen plagioclase is locally deformed resulting in the kinking of twin composition planes. Minor isolated brown biotite flakes exhibit variable alteration to chlorite and iron oxide. Intergranular phases comprise strained quartz, K-feldspar and microcline; magnetite, zircon and possible allanite are accessories.

Coilacriech Granite (LiG)

The unusual nature of the mica in granite to the south and west of Glen Gairn was first recognised by Tocher (1958). He thought it was phlogopite, and it was Hall and Walsh (1972) who first identified zinnwaldite at Gairnshiel in the Glenbuchat district. Harrison (1987), who considered the granite to be the last intrusive unit in his Glen Gairn Complex, also recorded 'sparse exposures to the south of Geallaig Hill' which he suggested 'indicate the presence of another body' of zinnwaldite granite. Lithium micas are now known to extend more widely (Webb et al., 1992) and it is their distribution which defines the Coilacriech Granite.

The Coilacriech Granite occupies an area of about 20 km2, the bulk of which lies within the Ballater district.

The intrusion can traced from the southern slopes of Tom Liath [NO 32 03] in the north to the area of lower Glen Girnock [NO 31 94], and extends in an east–west direction from Hill of Candacraig to Geallaig Hill. A small (less than 1 km2) detached portion of the Coilacriech Granite, corresponding to Harrison's (1987) zinnwaldite granite outcrop, extends south-westwards from Gairnshiel into the district [NJ 28 00].

Except in the Rinabaich–Dalraddie area where it abuts against Abergeldie granite and granodiorite, the Coilacriech Granite is largely contained by Dalradian metasedimentary rocks or, in the east, by the complex of metabasic intrusions. The southern margin of the intrusion follows the change in strike of the country rocks around the closure of the Camlet anticline. The full extent of the granite is masked by extensive roof pendants which also largely account for its irregular outcrop pattern. The most extensive of the roof pendants is up to 1.5 km wide and can be traced for over 3 km southwards from Torbeg [NJ 325 001]. The distribution of these pendants indicates that there was considerable variation in the elevation of the roof.

The granite contact with the host Dalradian rocks is sharp with little granite veining, except on the south slopes of Hill of Candacraig [NO 346 995] where loose blocks indicate that pink microgranite sheets, up to 0.6 m thick with many apophyses and with marginal and lensoid pegmatites, are abundant. One small exposure indicates that the sheets trend predominantly north-north-west, parallel to the main granite contact and an adjacent quartz–feldspar porphyry dyke. This dyke is cut by two generations of pegmatite veins (Plate 19) which are thought to be related to the Coilacriech Granite. Blue beryl occurs on a joint surface adjacent to the granite veins.

The Coilacriech Granite is clearly distinguished from the Glen Gairn Granite on the basis of field appearance and relationships, and petrography (Table 5). It is typically medium to coarse grained, equigranular to porphyritic with tabular plagioclase megacrysts. The occurrence of both pink K-feldspar and white plagioclase give the rock a pink aspect. Mica, largely comprising ferrous Li-mica, is anomalously dark in hand specimen. The intrusion also includes areas of finer grained granite which is porphyritic with prominent grey quartz and less abundant tabular plagioclase. Ferrous Li-mica plates of similar grain size to those in the bulk of the intrusion comprise up to 20 per cent of the rock, the remainder of which is fine grained quartzofeldspathic and pink in colour. Fine- to medium-grained granite exposed beside the Gairnshiel to Crathie road [NJ 2865 0032] contains irregular patches of pegmatite. In most instances it is not clear whether these fine-grained variants are related to unseen granite margins. However, on Creag Ghiubhais [NO 313 954] the finer grained granite which caps the typical medium to coarse Coilacriech Granite may well be part of the roof zone. The contact between the two rocks is visible 80 m below the summit [NO 3160 9543] where it is seen to be sharp and dipping gently to the south. Elsewhere, a heterogeneous marginal microgranitepegmatite facies of the granite is developed within about 50 m of the contact.

In thin section the plagioclase is tabular and unzoned, the K-feldspar is usually microcline-perthite and the mica has distinctive colourless to pale greyish brown or greyish green pleochroism. The mica is also marked by abundant deep greyish brown pleochroic haloes and in places colour zoning. It was described by Webb et al. (1992) as a Li-mica. 'True zinnwaldite' was described only from the vicinity of Gairnshiel in the Glenbuchat district. Li-mica specimens from close to the Deeside road east of Rinabaich [NO 307 965] and Creag na Creiche [NO 310 978] were reported by Harrison (1987) as containing about 41 to 42 per cent SiO2, about 16 per cent FeO and generally les than 1 per cent MgO. This gives Li2O = 2.3 per cent using the technique for estimating the Li2O of trioctahedral micas proposed by Tindle and Webb (1990).

Orthoclase encloses plagioclase, and quartz and micro-cline locally enclose perthitic orthoclase. Ferroan Li-mica occurs in the form of plates and irregular flakes which locally enclose plagioclase and are generally interstitial to plagioclase and quartz and to a lesser extent K-feldspar. In places, ferroan Li-mica is intergrown with K-feldspar in the form of irregular veins and patches. Vermicular inter-growths of quartz occur within Li-mica adjacent to some K-feldspar grain boundaries. Subsolidus grain boundary migration is suggested by the general occurrence of irregular sutured grain boundaries (see below). Other subsolidus features include both the replacement of plagioclase by topaz up to 1 mm across, and the occurrence of fluorite overgrowing ferroan Li-mica and plagioclase, particularly in the area to the north-west of the West Milton Burn [NO 306 998]. Plagioclase ranges from fresh to sericitised, whereas K-feldspar shows a brownish grey mottling thought to reflect kaolinisation. Fluorite is also present on joints.

Griesens

Patches of green fine-grained mica, mostly less than 1 cm in size and representing an incipient stage of pneumatolytic alteration (greisenisation) are developed sporadically within the western arm of the Coilacriech Granite to the north of the Cam Dearg–Creag na Creiche watershed. In places greisenisation is more extensive, with anastomosing veins of fine-grained green mica yielding a distinct greenish grey colour to silicified granite, as seen to the north-west of the West Milton Burn around [NO 3060 9983]. Ferroan Li-mica is particularly prominent in these altered rocks, occurring in plates and aggregates up to 6 mm in size. Mariolitic cavities are also present within the granite. A more advanced stage of greisenisation is seen 300 m to the south [NO 3056 9953], where the feldspars are largely replaced by quartz and green mica as described below.

In thin section, samples showing the early stages of greisenisation display a complex set of replacement reactions. Relicts of optically continuous plagioclase within K-feldspar result from the replacement of plagioclase by K-feldspar whereas K-feldspar is replaced by quartz in the form of a network of interconnected veins and patches. In places where orthoclase is rimmed by microcline, the latter may display such graphic texture whereas orthoclase is free of quartz. Radiating mats or irregular patches of fine-grained colourless or pale yellow mica replace feldspars (plagioclase preferentially) and occur as overgrowths on Li-micas. Quartz and mica growth are not always spatially related. Such textures are restricted to limited areas of some samples but are developed extensively in others. With further greisenisation and associated silicification, plagioclase is totally replaced, yielding a rock composed of quartz (which invariably contains abundant fluid inclusions), areas of fine-grained colourless to pale yellow mica and ferroan Li-mica. K-feldspar may occur as relicts within quartz. Several generations of quartz growth are shown by crosscutting vein structures. Fine-grained mica occurs as pseudomorphs after feldspar, as irregular mats or ferrous Li-mica overgrowths, or as thin veins that both cross-cut and are cross-cut by the quartz veins.

The granite is cut by quartz reefs, one of which is up to 10 m thick on Cam Dearg [NO 3179 9831], although more typically they are up to 1 m. The quartz veins contain vuggy cavities and form anastomosing networks which contain inclusions of greisenised granite on Cam Dearg. Quartz veins typically trend north-north-west around Cam Dearg and east-south-east to the north-west of the West Milton Burn. The full spatial extent of greisenisation is obscured by extensive till cover, although it appears to be restricted to the area to the north of the Cam Dearg–Creag na Creiche watershed. It is not clear whether an apparent correlation between the occurrence of anastomosing quartz veins and greisenisation is real or coincidental.

Ballater Granite (GB)

The Ballater Granite occupies an area of 30 km2 of the Ballater district which straddles the River Dee between Ballater and Cambus O' May (Figure 13); it also extends into the Glenbuchat, Alford and Aboyne districts.

The granite is generally well exposed, and consists of two lithotypes (Harrison, 1987). Thus the central and north-western parts of the pluton, in the Craigendarroch, Pass of Ballater and Crannach Hill areas, are composed of medium- to rather coarse-grained (3 to 5 mm) sparsely porphyritic to porphyritic granite with K-feldspar megacrysts up to 25 mm in size, whereas to the east, in the Vat Burn and Cnoc Dubh areas [NO 42 99] the granite is very coarse grained (5 to 20 mm) with K-feldspar megacrysts up to 50 mm in size. Gradations between medium to coarse and coarse-grained granite are seen to the north-northwest of Tomnakeist around [NO 399 990] where the grain size increases with increasing topographical elevation to the north. Contacts at [NO 3965 9974] between patches of medium-grained sparsely porphyritic granite and coarse-grained porphyritic granite are gradational over several centimetres. However, relationships with the very coarse granites in the south-east are not exposed.

Apart from grain size, there is little change in lithology across the pluton. The granite is typically pink. K-feldspar megacrysts and groundmass plagioclase are typically pale pink or white in colour. Quartz occurs as grey single crystals and aggregates which are particularly prominent as a result of their colour and grain size in the very coarse-grained variants. Biotite typically accounts for up to 5 per cent of the modal composition (Table 5).

In thin section, all granite samples show similar textural and mineralogical features irrespective of grain size. K-feldspar, mostly in the form of orthoclase, occurs in the form of large coarsely perthitic and poikilitic anhedra, some of which show patchy inversion to microcline. Orthoclase generally poikilitically encloses quartz or is interstitial with thin strips of orthoclase around some quartz grain boundaries. Lobate intergrowths of quartz and orthoclase indicate some overlap in the development of the two minerals. Tabular plagioclase crystals poikilitically enclosed by quartz and K-feldspar are always present. Many are around 1 mm in size and contain subhedral to euhedral cores wrapped by an irregular optically distinct rim with embayed boundaries against enclosing minerals. Some plagioclase is coarse grained (up to 5 mm long) with interstitial quartz and K-feldspar. Quartz, typically as large rounded grains or aggregates, is rarely included in plagioclase but is not uncommon in K-feldspar.

Two generations of biotite are recognised; in some rocks, medium-brown biotites largely altered to chlorite are included within all other major minerals. Accessory minerals, including apatite, opaque minerals, zircon and sphene, together with possible monazite in one sample, occur mostly within or adjacent to these biotites. In places such biotites form mafic schlieren along with tabular plagioclase. These are interpreted as restite material. The second generation of biotite is more abundant and occurs in all samples. It is very strongly pleochroic from pale smoky grey to very dark brown or greenish brown with strong pleochroic haloes and in places has chlorite interlayered along cleavages. The biotite occurs as late interstitial flakes, some of which anastomose along grain boundaries, and are intergrown with K-feldspar or, in one example, occurs in a fracture within K-feldspar. Locally these biotites are partially replaced by fluorite. They share several features with the ferrous Li-micas of the Coilacriech Granite including their mode of occurrence. Optically they fall between them and typical biotites. Locally, colourless to very pale brown micas partially replace plagioclase and to a lesser extent K-feldspar. They occur in the form of ragged interfingering flakes up to 1 mm in length. Some are interlayered with the late biotites.

Inclusions of medium-grained equigranular granite occur sporadically throughout the pluton. Typically, they are 15 cm across but may be up to 20 m. They are associated with local pegmatite in the form of segregations and bifurcating veins together with biotite-rich 'clots' and layers. Contact relationships are commonly ambiguous such that it is not always apparent whether they represent xenoliths. However, at several localities irregular patches and sheets of fine- to medium-grained granite up to 10 cm thick undoubtedly intrude the host granite. On Cnoc Dubh [NO 4222 9912] feldspar megacrysts project from the host granite into a near-vertical vein of medium-grained equigranular granite, 5 cm in thickness and trending 03°.

Pink quartz-phyric, fine- to medium-grained (1 to 2 mm) granite occurs as a north-trending dyke, 10 to 15 m wide, 2 km south-west of Culblean Hill [NO 3965 9974]. Grey quartz megacrysts and aggregates of grains are up to 6 mm in size. Feldspar megacrysts are less prominent, but are of a similar size. Apparent gradations between coarse porphyritic granite and quartz- and feldspar-phyric microgranite are recognised 1.2 km west-south-west of Culblean Hill. Quartz and feldspar megacrysts are comparable to groundmass minerals in the host granite but occur within a medium-grained quartzofeldspathic groundmass. The megacrysts comprise both monomineralic grains and granite xenocrysts; the latter are mineralogically similar to the groundmass although of different grain size.

The contact between the Ballater Granite and the host metasedimentary rocks is exposed in the Tullich Burn [NJ 3754 0066], immediately to the north of the district. Here, the contact is sharp and dips south-west at 15°; it is generally planar but steps vertically by a few centimetres across some joints, suggesting that emplacement was partly joint controlled. Rare microgranite-pegmatite veins occur within 10 m of the contact. A marginal facies of the granite, about 50 m wide, exhibits heterogeneous textures with variation from microgranite to pegmatite to coarse-grained granite over short distances. Patches of microgranite and pegmatite on Crannach Hill [NO 386 995] are interpreted as indicating proximity to the granite roof.

The southern and eastern margins do not seem to be xenolithic, and there is evidence that the regional country rock fabric has been steepened during intrusion in the area between the Ballater and Mount Battock granites [NO 40 94]. On the eastern margin both granite and county rock are shot through with felsite and aplitic microgranite. The presence of vuggy patches and coarse pegmatite within the intrusion is further evidence that the Ballater Granite was characterised by abundant late stage fluids, in contrast to the 'drier' Mount Battock and Lochnagar plutons. Part of the southern margin of the granite is fault bounded [NO 390 948] , with vuggy vein quartz and hematite along the broad north–south fault zone. The fault line extends for a kilometre northwards into the granite, giving rise to a zone, up to 30 m wide, of intense quartz veining together with epidotisation and hematitisation of the granite. Thin quartz veins are a feature of the granite to the west of the Tullich Burn; most are less than 10 cm in thickness, although they are up to 0.7 m on the south side of Craigendarroch [NO 3655 9639], and in places are spaced at 2 to 10 cm intervals. They are parallel to the principal joint set in the granite, being vertical with a strike of 170°. Drusy cavities are present in many of these veins.

A low-angle, irregular and bifurcating sheet of micro-granite, 8 to 20 m thick, is exposed in the crags on the south side of Craigendarroch, where it cuts the medium to coarse granite. The upper contact of the microgranite is near flat lying whereas north-eastwards at [NO 3670 9655] the lower contact is seen to dip north at 45° to 60°. Contacts between the microgranite and associated pegmatite with the host granite are not sharp, indicating near contemporaneity of the two intrusions. The micro-granite is composed of quartz, which has partly replaced the other components, K-feldspar (microcline and orthoclase), plagioclase, typically in unzoned tabular laths, and interstitial ferrous Li-mica. The microgranite is nearly equigranular with a grain size of 0.4 mm. The upper parts of the sheet display a subhorizontal pegmatite layering. Pegmatite layers are typically 2 cm thick with irregular spacing of typically 10 cm. Adjacent to the upper contact, pods and layers of pegmatite have a grain size up to 10 cm. Lithologically the microgranite is very similar to parts of the Coilacriech Granite.

Heddle (1901) reported zinnwaldite in association with muscovite, microcline, beryl, fluorite and quartz from the east end of the Pass of Ballater. A further occurrence at 'Monaltrie Cliff, reported to be also in the Pass of Ballater, occurs in 'druses within coarse graphic granite'. Bertrandite [Be4S12O7 (OH)2] recorded in the pink granite from the Pass of Ballater is one of only two known occurrences of this mineral in Scotland (Starkey, 1990).

Mount Battock Granite

This is the second largest pluton of the Cairngorm Suite, cropping out over 370 km2. However, only the westernmost 16 km2 of the outcrop lies within the Ballater district, the remainder lying in the Aboyne (66W) and Banchory (66E) districts. The pluton which is described by Gould (2001) comprises a number of different granite types of which only the Main Non-porphyritic Granite, the Cock Cairn Granite and some bodies of microgranite crop out in the Ballater district. Owing to poor exposure, cross-cutting relationships between phases are rarely seen, and the interpretation of emplacement order is based to a considerable extent on the outcrop pattern. However, the contacts within the Ballater district can be defined to within 25 to 150 m by means of solid exposure and blockfields. The western edge of the pluton is partly faulted, but for over half of its length it appears to be a normal intrusive contact.

Main Non-porphyritic Granite (GM1)

This phase forms the greatest part of the outcrop in the Ballater district. It is an equigranular, non-porphyritic granite with a typical grain size of 3 mm. It is typical primary-textured granite in the sense of Cobbing et al. (1992); it is believed to be one of the earlier phases to have crystallised. It is well exposed on the Cairn Nairvie crags around [NO 423 930] , on the crags at the top of Corrach corrie [NO 403 879] to [NO 407 876], and on the upper slopes of Mount Keen [NO 409 869].

All of the exposures show a certain amount of reddening, due to oxidation of iron in feldspars by circulating groundwater during the later stages of cooling, but between the Shiel of Glen Tanar [NO 401 895] and the end of the vehicular track [NO 400 886] there are abundant blocks of bleached granite, with pale green to white feldspars (suggesting alteration by hydrothermal fluids) and abundant blocks of breccia consisting of angular fragments of vein quartz cemented by later quartz.

Cock Cairn Granite (GM4)

This is the most mafic phase of the Mount Battock Granite. The largest outcrop of this phase extends into the Ballater district from the col separating Mount Keen and Braid Cairn [NO 418 867] to the southern contact of the pluton [NO 420 851]. There is also a small outcrop on the northern slopes of Mount Keen at [NO 410 875]. The shape of its outcrops suggests that the phase forms sheets or irregular bodies intruding the Main Non-porphyritic Granite.

The granite consists of a relatively fine-grained (1 to 1.5 mm) groundmass, with scattered K-feldspar crystals up to 8 mm and some quartz and biotite crystals up to 2 mm. Like the Main Non-porphyritic Granite it is primary-textured; the feldspar crystals are strongly zoned up to their margins and show no signs of disruption. Plagioclase and K-feldspar are equally abundant in the groundmass, and the phenocrysts are all orthoclase. Biotite forms 5 to 8 per cent of the rock; the accessory minerals are magnetite, apatite, zircon and sphene, together forming 1 per cent or less of the rock. The contacts with the M1 are diffuse in places, with possible patchy replacement of M1 by the Cock Cairn Granite.

Microgranite (FM)

Several bodies of microgranite occur in the district. Contacts with the M1 are sharp, and can be mapped to an accuracy of 5 m in places, for example on Cairn Nairvie [NO 420 931] to [NO 423 935]. The principal bodies of microgranite are: (i) Cairn Nairvie to the northern contact north-east of Black Moss; (ii) east and west of Etnach [NO 416 915] and (iii) Red Craig [NO 420 905]. These may be steeply inclined sheets or small stocks. Other, smaller, patches of microgranite are widely scattered. They occur not only as sheets and dykes, but also as irregular patches within the GM1 granite, especially near the margins.

Petrographically, the microgranite is a pink non-porphyritic rock with a grain size in the range 0.5 to 1.5 mm.

The texture is granular, hypidiomorphic, tending to idiomorphic, aplitic in the finer grained varieties, with drusy cavities in some of the more irregular patches. Open joints in the microgranite are, in places, lined with pegmatite.

Quartz veins

Three prominent en échelon quartz veins occur to the east and north of the Knowe of Crippley between [NO 399 658] and [NO 405 849]. They are 2 to 4 m wide, and each can be traced for up to 200 m. They consist of large blocks of vein quartz, which have been fractured along joints and then partly recemented by later quartz. For several metres adjacent to the veins, the country rock consists of microgranite, in part aplitic, which has been silicified, brecciated, and recemented by later quartz. Drusy pockets and red iron oxides (probably after pyrite) occur in this zone. The veins are orientated roughly parallel to the Knowe of Crippley Fault, and to the fault to the west of the Black Burn, and indicate a period of tensional stress shortly after emplacement.

Chapter 14 Post-tectonic major intrusions–evolution, age and emplacement

Geochemical characteristics

Various aspects of the geochemistry of the major granitoid intrusions and the associated diorites in the Ballater district have been described by Hall and Walsh (1972), Harrison (1987), Jarvis (1987), O' Brien (1985), Orridge (1961), Rennie (1983), Tindle and Webb (1989) and Webb and Brown (1984), and a large number of major and trace element analyses are available. These are supplemented by 29 new analyses presented in (Table 6) and (Table 7), mostly from the Abergeldie Diorite Complex but including representatives of the Lochnagar and Glen Gairn granites and the Juanjorge and Cul nan Gad diorites. The following is a brief summary of the main geochemical characters of the various intrusions.

The post-tectonic granitoids of the Ballater district form part of the Cairngorm Suite of pink granites of Stephens and Halliday (1984), whereas the diorites and granodiorites have similarities with their Argyll Suite (Gould, 1995). It is evident, however, that the diorites, granodiorites and granites are compositionally related, and all are interpreted as being consanguineous parts of the East Grampians batholith (Plant et al., 1990) which gravity data indicate underlies the region (Rollin, 1984).

Overall, the rocks are of calc-alkaline, I-type character, but a range of chemical characteristics is present. At one end of the range is the continuum of compositions which includes the various diorites, the Abergeldie Complex rocks and the Lochnagar granites; these appear to represent a simple magmatic fractionation series, in which even the most evolved of them (the L3 granite) still plots with all the others, firmly in the field of late- to post-tectonic intrusions on a Pearce et at (1984) Rb v Y+ Nb diagram. The most evolved rocks overall are those of the lithium-enriched Coilacriech Granite. They are somewhat similar to those of the Cairngorm granites and plot well outside the late- to post-tectonic field; they appear to have undergone late magmatic and/or hydrothermal changes that has resulted in them plotting away from the simple fractionation line. The other granites tend to lie between the two extremes, the Ballater and Mount Battock granites being more akin to Coilacriech, and the Glen Gairn granites being more similar to those of Lochnagar. The differences between the various bodies are, to some extent, well displayed in the coloured plots of the stream sediment geochemistry displayed in the East Grampians Regional Geochemical Atlas (British Geological Survey, 1994), particularly in the plots for Be, Li, Rb, Sr and U.

In common with Caledonian post-tectonic intrusive rocks occurring elsewhere in Scotland, these rocks exhibit evolving trends towards enrichment in SiO2, K2O, and Rb and depletion in TiO2, MnO, MgO, CaO, P2O5 and Sr. Harker plots of major oxides and trace elements against SiO2 (Figure 21) show, in the case of the diorites and the early Lochnagar and Glen Gairn granites, smooth trends with little scatter of points, and overlap of fields between the various intrusions indicating a simple fractionation path from a common parent magma. The diagrams highlight the wide range of chemical composition of these rocks, particularly those of the Abergeldie Complex.

MgO is rather more variable than other oxides among the diorites and granodiorites and it is possible to detect two broad trends with a compositional difference of about 1 per cent MgO, although these tend to merge in rocks with more than 62 per cent SiO2. The high-MgO series reflects the common occurrence of partly assimilated cognate xenolithic material which appears as fine-grained mafic spots and blotches in some rocks.

The contrast between the Juanjorge and Glen Doll diorites which is so evident in their petrography (Chapter 12) is also very noticeable in their geochemistry. For example, diorites of Glen Doll are enriched in FeO relative to the other intrusions. The Glen Doll diorites are also enriched in zinc and copper relative to the other rocks. Zinc concentrations are in the range 75 to 137 ppm compared with 41 to 89 ppm and 35 to ppm respectively in the Juanjorge and Abergeldie rocks. Comparable figures for Cu are 25 to 120, 5 to 42 and 9 to 26 ppm respectively. Jarvis (1987) concluded that fractional crystallisation could not produce the enhanced zinc (and presumably copper) values, believing instead that they resulted from contamination of the magma by Argyll Group metasedimentary rocks which to the southwest are known to carry stratabound zinc and copper sulphide mineralisation (Smith et al., 1984). Contamination by such rocks would also account for the strong iron enrichment as the stratabound mineralisation is typified by high pyrite concentrations. The Juanjorge dioritic rocks, despite their differences from the Glen Doll diorites, show similar major and trace element concentrations to the Glen Doll monzogranite phase although the REEs have contrasting patterns (Jarvis, 1987).

The diorites of Juanjorge, Cul nan Gad and the Abergeldie Complex are very similar in composition and lie on simple fractionation trends which pass through the Abergeldie granodiorites and monzogranites to the Lochnagar and Glen Gairn granites (Figure 21); the most evolved phases are the Lochnagar L3 and Glen Gairn G2 granites.

There is an overall similarity in the compositional range of the Lochnagar and Glen Gairn granites, although there is no simple correlation between them. The less evolved group of Glen Gairn granites, GG, overlaps the field of Lochnagar L1, and at least part of L2. The L1 and L2 fields also overlap with one another, although the latter generally has higher SiO2 and K2O, and lower Sr. The most evolved of the Lochnagar granites, L3, continues this trend with 76 to 79 per cent SiO2 and generally lower Sr, in some cases less than 10 ppm Sr. Their composition is somewhat similar to that of the pink granites of Ballater, Coilacriech, Mount Battock (and Cairngorm) but, although Rb is enriched, it is significantly lower (mostly < 200 ppm) in L3 than in those granites (where it is typically > 400 ppm). Nb and Y in L3 are also characteristically lower (< 20 ppm) than in the other pink granites, as are Th (< 20 ppm) and U (< 10 ppm). There are no signs of Li enrichment among the Lochnagar analyses of O'Brien (1985), although Tindle (personal communication, 1991) estimated Li2O contents of 0.5 to 1.0 per cent in biotites from L3 granite (compare with 2.3 per cent Li2O in micas from the Coilacriech Granite). The only other sign of Li enrichment in the Lochnagar pluton is a report by Riddler (1967) of lepidolite (with epidote and kaolin) at the sheared and brecciated margin of the L1 granite near the north-west end of Loch Callater in the Braemar district (Sheet 65W). The pink evolved Glen Gairn Granite, iGG, is similar to L3 in composition as well as in mineralogy and texture, as are the Khantore and Gourock granites. Mount Battock, Ballater and Coilacriech are all evolved granites, with SiO2 similar to Lochnagar L2 and L3. However, they are significantly different in composition, particularly in various trace elements.

A full description of the Mount Battock granites and their geochemical characteristics is provided by Gould (2001). Two analyses of the portion lying in the Ballater district are given by Webb and Brown (1984). The first, representing the Main Non-porphyritic Granite, GMi, has a similar SiO2 content to the Lochnagar L2 Granite but contents of Rb, Y, Nb, Th and U are all at least 50 per cent higher. The second, a microgranite from the same locality, is somewhat similar to the Coilacriech Granite, with higher Rb (500 ppm), Y (73 ppm), Nb (47 ppm) and U (27 ppm). The lower Th content (17 ppm) may reflect the influence of volatiles. The Ballater granites are generally similar chemically to Mount Battock, the analyses in Webb and Brown (1984) indicating that they are all more evolved than the L2 granites of Lochnagar, with Rb ranging from 233 to 497 ppm and with low Sr. In contrast to the GM1 Mount Battock Granite sample quoted above, Th (17 to 71 ppm) is always more abundant than U (4 to 22 ppm). Within the Ballater Granite the coarse-grained porphyritic variety forming the south-eastern part is less evolved than medium-grained and more equigranular granite in the west and north-west. The latter displays more evidence of late- to post-magmatic processes in the form of fluorite and quartz reefs.

Coilacriech is the most evolved and geochemically distinctive of all the granites in the Ballater district. This is most evident in the very high Rb (500 to 1200 ppm) and very low Sr (< 20 ppm), but a number of other elements are also anomalous, notably Li (511 ppm), Y (96 ppm) and Nb (116 ppm). There are additionally local concentrations of Sn (119 ppm) and Mo (57 ppm) (O'Brien, 1985). The granite is apparently more evolved towards the north, where there is local development of fluorite and topaz, and to the north-west where maximum Li enrichment is associated with the separate boss of zinnwaldite granite at Gairnshiel Bridge. Here, there are thin greisen and aplitic veins in the granite, and W–Sn–Mo–Bi–Ag mineralisation with quartz and pyrite has been recorded (Webb et al., 1992).

Evolution

The numerous petrographical, whole rock and mineral chemistry studies on these rocks, most of which have been cited previously, have helped to elucidate their origin and evolution. In addition, stable and radiogenic isotope research by Harrison (1987) and Jarvis (1987), upon which much of this section is based, have provided some pointers to the possible origins of the magma.

In the Glen Doll Diorite, initial 87Sr/86Sr ratios display a wide range of values within most rock types (0.70632 to 0.70836 in diorites and 0.70653 to 0.71137 in monzogranites), suggesting that the initial mantle-derived magma was heavily contaminated with more radiogenic upper crustal material (Jarvis, 1987). The isotopic data also confirm the field evidence that the monzogranites, diorite and gabbros are separate intrusive phases. Both basic and intermediate rocks, as evidenced by wall and floor-parallel layering, were produced by cumulate processes, which are more likely to have resulted from local rather than large-scale convective fractionation. The presence of discrete pyroxene- and hornblende-rich units in the intercumulus phase indicates that the water content of the magma was quite variable. The development of coarse appinite in the complex is evidence of late migration of interstitial fluids enriched in volatiles, which in their upward movement locally disrupted and destroyed the igneous fabric. Field relationships support the isotopic evidence that after the formation of the cumulates there was widespread contamination of the magma by partially assimilated Dalradian metasediments, leading to enhanced Fe, Cu and Zn levels in the dioritic rocks. Further contamination of the partly solidified diorite magma by granitic magma, generated by partial melting of xenoliths, led to the formation of quartz-monzodiorite and granodiorite in the south-east part of the intrusion. Field and trace element data clearly indicate that the marginal monzogranite was also formed from partial melting of Dalradian xenoliths by the diorite magma.

In contrast to Glen Doll, the other dioritic bodies are similar in many respects to late Caledonian plutons elsewhere in Scotland. For example, the range of lithologies is comparable to that shown by the Cruachan Complex. In addition, the geochemical trends shown by these rocks definitely reflect an origin by processes of fractional crystallisation with little evidence of crustal contamination. Jarvis (1987) noted that the bulk rock compositions of the Juanjorge rocks fall on the liquid line of descent shown by the early Devonian calc-alkaline lavas and as such probably reflect original magma composition. It is considered likely that the diorite to early granite suite was derived by fractionation of a single magma, although magma mixing cannot be entirely ruled out.

Harrison and Hutchison (1987) subdivided a number of the Newer Granites into an older (about 415 Ma) and younger (about 408 Ma) suite. They suggested that the older intrusions, which include Lochnagar, Glen Gairn GG and Mount Battock, were not mantle derived, but instead formed by subduction-initiated melting of lower crust. This conclusion is based on the predominance of hydrous mafic silicate phases (hornblende and biotite) and the low concentrations of Ni in the diorites. Glen Gairn iGG and Coilacriech form part of the younger suite, all of which are, apparently, spatially related to the intersections of major lineaments and north-east-trending faults. Harrison (1987) suggested that these granites postdated active subduction, and believed them to have been generated by partial melting of the earlier suite in a process initiated by fluxing mantle-derived volatiles, which were concentrated in these areas of structural weakness. This general picture must be modified by the recognition that the Lochnagar L3 granite belongs to the younger suite.

Chronology

Early U–Pb dating of zircons by Pidgeon and Aftalion (1978) established that the Lochnagar granites were emplaced between 400 and 415 Ma. This was apparently confirmed by subsequent Rb/Sr ages of 415 ± 5 Ma for Lochnagar (Halliday et al., 1979) and 404 ± 6 Ma for Glen Gairn iGG (Harrison and Hutchison, 1987), but as both of these determinations were made on combinations of rocks from more than one intrusive phase their reliability is questionable.

Three new U–Pb age determinations from the Ballater district have been obtained from samples of Abergeldie Complex diorite and Lochnagar L2 and L3 granites. The diorite recorded an age of 423.4 ± 2.7 Ma based upon analysis of one titanite and three zircons. A broadly comparable age of 426.2 ± 2.5 Ma was obtained from analysis of five zircons in the L2 granite. The L3 granite sample, however, yielded an age of 417 ± 1 Ma from two monazites. The diorite date is recognised as the age of the oldest phase in the Abergeldie Complex and, because of the close spatial, petrological and chemical relationship of these rocks to both the Lochnagar and Glen Gairn granites, it might also be regarded as marking the onset of their development.

The L2 and L3 granites are seen to cut L1 granite which is demonstrably younger than the dated Abergeldie Complex rock. Hence, the order of intrusion of the Lochnagar granites is now thought to be L1–microgranite–L2–microgranite–L3, somewhat different to that proposed by Oldershaw (1974). Goodman and Lappin (1996) have indicated that the Cul nan Gad and Allt Darrarie intrusions are older than the Juanjorge body. The presence of a flattening fabric in parts of the Abergeldie, Cul nan Gad and Allt Darrarie bodies together with the complex nature of the Juanjorge–Li contact would seem to suggest that the diorite–granodiorite suite was emplaced over a period of time, but only shortly before intrusion of the L1 granite.

Glen Gairn GG granite is petrologically and chemically similar to L1 and they may be coeval. Harrison (1987) noted that the emplacement of the two Glen Gairn granites was separated in time by the intrusion of microdiorite and calc-alkaline lamprophyre dykes. L3 and Glen Gairn G2, are chemically very similar but as L3 is cut by microdiorite [NO 2622 9225] it seems unlikely that the two granites are coeval.

No radiometric age determinations have been carried out on the Ballater Granite. According to Webb and Brown (1984) the western part is the more evolved and hence may be younger than the eastern part. The enhanced lithium content, and the geographical proximity suggests some affinity with the Coilacriech Granite, but the age relation between the two is not known. Harrison and Hutchison (1987) obtained an Rb/Sr whole rock date of 416 ± 4 Ma for the Mount Battock Main Granite.

Emplacement

The mode of emplacement of the Newer Granites was divided by Read (1961) into forceful or active intrusions, where space is created by pushing aside the country rocks and insitu 'ballooning', and permitted or passive intrusions where emplacement is allowed by structurally controlled subsidence of roofing rocks. Read's permitted granitoids, in the strict sense, comprise only the high level (subvolcanic) ones whose emplacement was most likely connected with caldera formation. Permitted granites, in a broader interpretation, encompass bodies which were also emplaced by gentle stoping and assimilation of country rock. With this caveat it can be said that the late Caledonian major intrusions in the Ballater district include examples of both forceful and permitted intrusions.

For example, fabrics in the country rock adjacent to the western margin of the Mount Battock Granite suggest that the granite was forcefully emplaced. In particular, the foliation within the psammites, which regionally dips at 30° to 40° to the north-west steepens to 70° to 80° adjacent to the granite margin. Field evidence which relates the steepening to granite emplacement includes strong downdip lineations and slickensides, and shear fabrics, although in more pelitic lithologies the latter may in places be difficult to distinguish from the already strong regional foliation. In psammitic units, however, rotated feldspar porphyroclasts may be visible in hand specimen, along with incipient C–S fabrics (Lister and Snoke, 1984); more distinctive still are the anastomosing bands of mylonite and cataclasite. The evidence shows that the steepening was accomplished by shearing with a thrust sense parallel to the regional foliation, thus causing an overall shortening across the gneisses as the granite was emplaced. It is possible that there was a regional-scale shear zone crossing the Ballater district prior to granite emplacement (compare with Ashcroft et al., 1984), a line of weakness exploited by the granite on intrusion. However, the lateral restriction of the shear fabrics and steepening towards the area close to the granite margin, suggests that forceful emplacement of the granite is a more likely cause. It has been suggested that the shape of metamorphic isograds to the north of the Mount Battock Granite may be due to doming during granite intrusion (Chinner and Heseltine, 1979). There is some evidence, in the north-west of the granite body, for the flattening deformation affecting the outer parts of the pluton itself, with fabrics defined by the orientation of potassium feldspars and fluid inclusions in quartz, paralleling its margins (information from T N Harrison, personal communication, 1987). All the shear fabrics in the psammites are now annealed to some degree, and there is a progression from ductile mylonites through to brittle cataclasites, indicating a long timescale for deformation, as the rocks became progressively cooler.

The two Glen Gairn and the Coliacriech granites appear to have been largely permitted intrusions, the dominant process in the emplacement being stoping of the country rock (Harrison, 1987). However, there may well also have been a forceful element in the intrusion of the Coilacriech Granite which caused the opening of the Camlet Anticline (Chapter 10). The roof zone of Glen Gairn Granite appears to have been predominantly a vein complex, a large portion of which survives to the north and west of Geallaig Hill. There is an easterly increase in the volume of included country rock, suggesting that the roof sloped in that direction.

It was the shape of the Lochnagar Complex outcrop which prompted MacGregor (1948) to suggest that it 'may possess some kind of annular or ring form similar to that presented by the Etive granite-complex'. Although Read (1961) subsequently admitted that forcefully emplaced intrusions such as Ardara in Co Donegal have comparable aspects, Oldershaw (1974) perpetuated the ring complex interpretation maintaining that L2, L3 and two central microgranites were emplaced by successive cauldron subsidence. Oldershaw's conclusions were questioned by Rennie (1983) who noted that none of the features normally associated with ring complexes, le multiple intrusion of ring dykes, flinty-crush along ring faults, xenolith-filled dykes and breccia pipes', was present in Lochnagar. To this list might be added the absence of any evidence of contemporaneous volcanicity in the Ballater district. Modern research (Goodman and Lappin, 1996) indicates that, rather than being a high-level subvolcanic intrusion, Lochnagar was intruded under a confining pressure of 2 to 3 kbar, equivalent to a depth of 7 to 8 km. There is limited evidence that the L1 granite was emplaced forcefully; in places there is a common orientation of discoidal cognate xenoliths and some preferred orientation of the feldspar megacrysts parallel to the granite margin (Oldershaw, 1974). There is more compelling evidence (including for example, flattening of fringing dioritic and granodioritic rocks and slight reorientation of the country rocks) that the L1 granite was actively emplaced, but not as forcefully as the Mount Battock Granite. The L2 and L3 granites and the microgranites are all unfoliated and there is little doubt that they were emplaced passively, as were the diorites and granodiorites which are largely unfoliated. The roof zones of all Lochnagar granites, except L2, and some of the diorites, are still preserved.

Thermal metamorphism

All of the late-tectonic and post-tectonic major intrusions caused some thermal overprint on the regional metamorphic assemblages in the Dalradian rocks. The aureoles around even quite small diorite bodies are generally more extensive than those marginal to even the largest granitic intrusions, pointing to the higher intrusion temperature of the former. This has been documented particularly for the diorites at the south-eastern margin of the Lochnagar granites (Goodman et al., 1990; Goodman and Lappin, 1996).

In the aureoles of the Lochnagar and Glen Doll/ Moulzie Burn intrusions, the rocks show a range of susceptibilities up to 10 SI, but they are never as magnetic as the adjacent diorite, quartz-diorite or tonalite, so that the external contact of the intrusions can always be clearly distinguished.

The following descriptions of textures, petrography, extent and conditions of thermal metamorphism is based in part on Phillips (1992a, b) and Goodman et al. (1990).

Thermal textures and petrology

Semipelite and pelite

Macroscopically, the effects of thermal metamorphism are most obvious in these lithologies since it invariably produces cordierite which imparts a distinctive blue-grey hue to the rocks. The rocks additionally loose their characteristic fissility, although compositional layering in semipelites is largely preserved. Reactions producing biotite and cordierite are initiated at the lower end of the thermal spectrum so their development is particularly widespread.

Cordierite occurs in a variety of forms ranging from granoblastic intergrowths to rounded or elongate porphyroblasts; these largely reflect the nature of the protolith. The porphyroblasts invariably contain tiny rounded inclusions of foxy red-brown biotite ((Plate 20)a) indicating that the growth of the two minerals was roughly coeval. Much of the cordierite is intergrown with biotite in the more pelitic layers, where it has developed by a reaction involving muscovite, eastonitic biotite and quartz, to the extent that primary muscovite is seldom if ever seen in these hornfelses. In this environment the outline of the cordierite seldom resembles that of the precursor, although in thin section (S82067) elongate grains are apparently mimicking earlier phyllosilicates ((Plate 20)b) in crenulated microlithons. Growth at the expense of garnet is more obvious, since the porphyroblastic form is retained and vestiges of the garnet are present in most cases. With progressive cordierite growth, largely intact garnet with a corona of granoblastic cordierite, feldspar and minor biotite is replaced by a symplectite of cordierite and biotite, which locally retains relict garnet. In layered semipelitic lithologies cordierite occurs preferentially within the more pelitic component, but it is also found growing across psammitic layers. Porphyroblast growth may be controlled by pre-existing structures, particularly the regional foliation. However, in one rock (S94395) a number of cordierites, up to 3.3 mm long are aligned along an incipient cross-cleavage at an angle of 40° to the foliation.

The occurrence of andalusite is more restricted than that of cordierite, being confined to pelitic layers, where its growth was controlled by the relative abundance of alumina. However, in rocks with regional sillimanite, growth of andalusite was not simply an isochemical conversion, but instead took the form of finely divided new growth, often as the product of feldspar breakdown. There are also some examples where mechanical comminution of feldspars has led to easier reaction and the formation of andalusite at some distance to the relevant intrusion. In rocks of low to moderate grade, the andalusite is usually present as porphyroblasts which may have a square or stubby prismatic shape, or more rarely as columnar or granular aggregates. The porphyroblasts are speckled with inclusions of regional and early thermal minerals, most notably foliation-parallel rod-like iron oxide trails and rounded foxy red-brown biotite.

With increased grade, thermal andalusite in some of the hornfelses is partly or entirely replaced by sillimanite, which is generally more prismatic than the fibrolitic regional sillimanite, although the fibrous variety also occurs. Unlike the growth of andalusite described above this is an isochemical conversion. Thus, in some of these higher grade rocks it is not uncommon to find square pseudomorphs of sillimanite replacing andalusite. These pseudomorphs are unstable against biotite, however, and are replaced by cordierite and a green spinel, hercynite ((Plate 20)e).

At the highest thermal grades encountered in the Ballater district the pelitic rocks consist largely of cordierite, sanidine, and hercynite, although sillimanite may also survive. These rocks are typically blue-black massive hornfelses, and in places a foliation defined by quartzofeldspathic layers survives from the original rock.

In some aluminous hornfelses the andalusite is not replaced by sillimanite or hercynite, but instead persists to high grade as rather ragged porphyroblasts. These rocks do not contain as much cordierite as the hercynite-bearing types, and biotite is preserved even at high grade. As a result, the hornfelses still show the regional foliation, though it is overprinted by the hornfels assemblage. In places this type of hornfels contains hypersthene, commonly with an elongate, cleaved appearance, associated with blue corundum and sanidine. The paucity of cordierite and persistence of andalusite to high grade when compared with hercynite-bearing assemblages is considered to be a function of oxidation ratio of the original gneisses (Chinner, 1960).

Contact melting is locally evident in areas of high-grade thermal metamorphism, most notably on the south-east side of the Glen Doll Diorite, where field evidence clearly demonstrates a link on all scales between the distribution of xenoliths and contamination of the quartz-diorite melt with a granitic melt. On a small scale there is a close association between the abundance of orthoclase within the igneous rocks and proximity to xenoliths. Xenoliths show evidence of partial melting in both the interfingering granitic segregations and the development of graphic textures within the segregations.

More limited contact melting, in a small enclave of semipelite within diorite of the Abergeldie Complex, is exposed in the wooded area 200 m west of Buailteach [NO 2753 9325]. The semipelite (S81978) has a pseudobrecciated texture resulting from the flowing of the quartzofeldspathic material between cordierite-rich boudins. The cordierite porphyroblasts and poikiloblasts may exceed 10 mm in length, and are notable for their volume of included corundum. The quartzofeldspathic component comprises a coarse-grained subhedral aggregate of quartz, K-feldspar and foxy red-brown biotite with subordinate plagioclase and muscovite. It is granitic in character, but the inclusion of corundum, the similarity of the biotite to that in the rest of the rock and the presence in places of marginal biotite selvedges point to an origin by localised partial melting rather than intrusion.

Some lensoid quartz–K-feldspar segregations 750 m east of Braenaloin [NO 2876 9997] may also reflect the onset of partial melting.

Calcsilicate rocks

Most of the impure limestones within the Appin Group succession which occur in the Crathie area lie sufficiently close to units of the Abergeldie Complex for talc-silicate hornfelses to develop. Typically they comprise a granoblastic aggregate of diopside and turbid to heavily sericitised plagioclase along with grossular garnet and sphene. Some free carbonate may be present and in places may constitute a major component. Hornblende, tremolite and biotite are less common, and where present have obviously grown at the expense of diopside. It is recognised that these mineral assemblages could result from regional metamorphism. However, the non-directional textures and the additional presence in some rocks of idocrase, scapolite, wollastonite (Heddle, 1901) and corundum are suggestive of a thermal overprint.

The occurrence of Argyll Group calcareous rocks within the aureole of the major intrusions is confined to the areas around Camlet–Bovaglie [NO 30 92] to [NO 30 93], The Knock [NO 3495 3595] and Pollagach Burn [NO 40 93] to [NO 41 95]. In the first named area, thin and generally discontinuous calcsilicate units within the Glen Girnock Calcareous Formation lie close to the roof of the Khantore Granite. In contrast to the Appin Group hornfelses, these have little free carbonate, but have significantly more amphibole, either in the form of hornblende or tremolite, although diopside and plagioclase remain the main phases. The additional presence in places of cordierite and red-brown biotite confirm that these are thermal assemblages.

In Pollagach Burn, coarse-grained calc-silicate hornfels is developed in rocks of the Water of Tanar Limestone Formation directly adjacent to the margin of the Ballater Granite. Many different assemblages have been described from these hornfelses (Hutchinson, 1933), the major minerals being diopside, wollastonite, grossular garnet and vesuvianite, with highly altered feldspar and possibly some scapolite and tremolite. There is little free calcite remaining in the rock, much having been used in reactions to form the new mineral suite. The assemblages are, however, typical of those produced in skarns and it may be that the formation of the talc-silicate hornfelses is not isochemical, but is metasomatic, with fluids introduced from the crystallising granite.

Metabasic rocks

The metabasic rocks, which contain the simple assemblage hornblende, plagioclase, quartz, sphene and opaque minerals, appear to be largely unaffected by low-and moderate-grade thermal metamorphism. However, those containing garnet show evidence of thermally induced change up to approximately 2 km away from the margin of the intrusions, the mineral becoming unstable against hornblende, and reacting to give a symplectite of hornblende, plagioclase and opaque minerals. Relict pieces of garnet remain within symplectite which may retain the outline of the original garnet. The regional assemblages which contain clinopyroxene usually also contain garnet, but even where the garnet has reacted out by the above reaction the clinopyroxene remains unaffected by the thermal metamorphism.

At highest grade the metabasic assemblages which do not contain garnet also become unstable. The hornblende begins to break down, blebs of labradoritic plagioclase develop along the cleavage planes and diopside and enstatite are formed. The product minerals form a fine equigranular mosaic, though the outlines of the original hornblende grains may be visible in the mismatch between the orientation of different domains within this mosaic. The resulting rock is a dark, green-black glassy hornfels, which is generally strongly jointed and with accessory minerals such as sphene or sulphides remobilised onto the joint faces.

Post-Caledonian intrusive rocks

Symptoms of thermal metamorphism in these rocks are considerably more localised and harder to quantify than in the Dalradian country rocks, but are important components in unravelling the intrusive history. Probably the most convincing evidence is to be found in the early fine-grained diorite phase of the Abergeldie Complex, particularly where it is present as small bodies within later phases. In these rocks (for example (S77614), (S77617) and (S77677)) the mafic phenocrysts have been pseudomorphed by clusters of pale green hornblende which in places have been partially replaced by red or red-brown biotite. The extent of recrystallisation in the groundmass is more variable; for example in (S77614) it comprises a fine granoblastic intergrowth of plagioclase, hornblende, biotite and finely divided opaque minerals, whereas in (S77677), which has a similar petrology, the plagioclase retains its tabular form and, along with hornblende and biotite, has a pronounced common alignment. (S77614) has some plagioclase megacrysts which may be zoned, but invariably contain granoblastic hornblende and apatite grains. These are thought to be related to the millimetre-scale vein of quartz-diorite which cuts the rock, and are not considered to be phenocrysts.

Jarvis (1987) considered that the dominance of biotite over hornblende and the reddening of feldspars in marginal areas of the Juanjorge Diorite represented the thermal effects of the Lochnagar Granite. There is also microscopic evidence of a thermal event in the hypersthene gabbro phase of Juanjorge with hornblende being replaced by red biotite. The development of red biotite and cloudy apatite in the Abergeldie Complex white granite and, more rarely, in the surrounding granodiorite, might also be interpreted as a thermal overprint, although the biotite shows no evidence of recrystallisation.

Distribution of thermal metamorphism

The known extent and degree of the thermal metamorphism surrounding the major intrusions, is illustrated in (Figure 17). On the basis of field and microscopic data the hornfelses have been divided into low to moderate grade, characterised by cordierite and andalusite, and high grade where sillimanite is the dominant aluminosilicate, and corundum and spinel are also present. Because of the patchy nature of the evidence these divisions are very generalised and no attempt has been made to quantify the outer limits of the aureoles.

The areal extent and intensity of thermal metamorphism are largely a feature of temperature and as such relate to the composition of the intrusion and the spatial relationship of the country rocks to the body. In effect, the highest grade hornfelses develop next to diorites rather than granites, and the aureole will be more widespread in the country rocks that form the roof than those of the side margins. The width of the aureole may also relate to the composition of the country rock; for instance, aluminous pelites will react at a lower temperature than their non-aluminous counterparts.

In the Ballater district, high-grade (pyroxene hornfels facies) thermal metamorphism is restricted to rocks surrounding dioritic bodies. The most extensive development is on the south-east side of the Lochnagar complex where, as a result of the overlapping Glendoll, Moulzie Burn and Juanjorge aureoles the high-grade hornfelses cover several square kilometres. Within the area there is clear evidence that thermal andalusite was replaced by symplectic aggregates of K-feldspar, cordierite, spinel, corundum and sillimanite. Rennie (1983) interpreted this as indicative of two thermal events which he attributed to successive emplacement of the Lochnagar L1 and L2 granites. However, taking into account the distance from L2 (3 to 4 km), the fact that the two granites have near-vertical margins and were emplaced synchronously, and the absence of high-grade hornfelses around other granites in the district, it seems more likely that this represents a single progressive event associated with diorites.

The Glen Doll aureole is 200 to 400 m wide around the south-east part of the intrusion. Ashcroft (1958) considered that the onset of thermal metamorphism was marked by a reddening of otherwise grey rocks. In the outer part of the aureole, the regional assemblage of biotite, garnet, kyanite and sillimanite has been replaced by thermal biotite, muscovite and cordierite with andalusite appearing nearer the intrusion (Ashcroft, 1958). The inner or high-grade part of the aureole is around 150 m wide in the Red Craig area and is characterised by the assemblage cordierite, sillimanite ((Plate 20)d) and K-feldspar with minor hercynite and corundum. The total absence of andalusite in the inner aureole is not mirrored in any other of the district's aureoles and may reflect the hotter nature of the Glen Doll intrusion and/or the composition of the country rock. In Winter Corrie on the south side of Glen Clova, the aureole is at least 200 m wide on the evidence of cordierite porphyroblasts, but in the Burn of Kilbo area little evidence of hornfelsing is apparent within 100 m of the contact. The sudden increase in width of the inner aureole on The Scorrie [NO 27 75] results from the rocks being on the roof of the diorite.

It is evident from (Figure 17) that on the east side of Lochnagar the width of the aureole and the development of high-grade hornfelses is spatially related to dioritic rocks. Thus the width of the aureole ranges from 300 m next to the granite north of Allt Darrarie, to 1.4 and 1.1 km, respectively, adjacent to the Moulzie Burn and Cul nan Gad diorites.

Thermal metamorphism is invariably present in Dalradian rocks north and west of Glen Girnock, but because of the complexity of the intrusive history, especially the areal overlap of many intrusions, it is difficult to quantify the effects of individual bodies. Also, because of the high proportion of intrusions relative to country rock the extent of aureoles cannot readily be defined. The bulk of the hornfelsing is of low to moderate grade, characterised by cordierite, red-brown biotite and minor andalusite, and probably relates to the Lochnagar, Glen Gairn and Khantore granites. However, these rocks locally contain assemblages and textures which are indicative of higher temperatures than would be expected with granites, for example corundum and hercynite in the aureole of the Khantore Granite ((Plate 20)e). The occurrence of prismatic sillimanite [NO 2655 9526] and partial melt textures [NO 2876 9997] and [NO 2753 9325] can be readily explained by their proximity to diorite, and the subsequent partial replacement of the sillimanite by andalusite ((Plate 20)f) is evidence that there was a second, retrogressive event presumably related to the emplacement of later granite or granodiorite phases of the complex. The localised presence of sillimanite in metasedimentary rocks on the roof of the Coilacriech Granite for example at [NO 3392 9803] is harder to explain since this particular intrusion is considered to have been one of the lowest temperature granites in the area. The mineral is intergrown with thermal biotite and andalusite so cannot be the product of regional metamorphism. However, whether it was generated by a concealed diorite body (for which there is no geophysical evidence) or by the adjacent late tectonic basic sheets is not at all obvious. With regard to the second alternative, there is evidence from a rusty weathering, pyritic, hornfelsed, gneissose semipelite [NO 2488 7706], on the north-east flank of Glen Doll, that an early thermal assemblage, attributable to the enclosing early tectonic amphibolite sheet, has survived subsequent kyanite and sillimanite grade regional metamorphism.

The aureole on the south-east side of the Coilacriech Granite is less than 200 m wide, but round the southern margin increases to over 1 km and coalesces with the Khantore Granite aureole. The increase in width probably results from one or both of the intrusions having gently inclined roofs. Some variation in aureole width is apparent round the Ballater Granite with up to 500 m on the south side but less than 100 m apparent in the Pollagach Gap to the east. For the most part hornfelsing by the Mount Battock Granite is confined to the initial 500 m west of the contact, but close to the Ballater Granite the combined aureole extends outwards for over 1 km.

Within the Ballater district there are a number of areas where low- to moderate-grade thermal metamorphism is apparent in Dalradian rocks but no intrusive rocks are exposed. The most extensive of these can be traced south-westwards from the Creag Liath area [NO 333 939] to Craig Megen, where it merges with the Cul nan Gad aureole and affects rocks of the Glen Girnock Calcareous Formation, especially the Craig Liath Pelite Member, the Meall Dubh Metabasite Formation and the Queen's Hill Gneiss Formation. The zone may link up with an area of hornfelsing, detectable only in aluminous pelites on the east side of Glen Muick (Figure 17). It is possible that the thermal event could be linked with the emplacement of syntectonic basic intrusions; certainly there is a thermal aureole round the Glen Mark metadiorite. However, the distribution of these rocks is far more restricted than that of the hornfelses, and the grade of metamorphism is not commensurate with such basic to ultrabasic rocks.

Another area of enigmatic thermal metamorphism is on the north side of Glen Doll between Craigs of Loch Esk [NO 242 788] and Cairn Lunkard [NO 234 781] where, in gneissose pelite, andalusite and cordierite overgrow sillimanite and kyanite relicts. The occurrence appears to be largely separated, by approximately 750 m of unaltered gneiss, from the narrow thermal aureole on the southern edge of the Lochnagar Granite, although a tongue of hornfels extends eastward from Loch Esk to coalesce with the Lochnagar aureole above West Corrie. The hornfelsing may relate to a granodiorite or diorite at depth or possibly an unexposed body in the Loch Esk area, where morainic material is abundant. The small granodiorite 'plutons' or lensoid masses mapped during this survey in Glen Doll suggest that other diorite or granodiorite masses may lie just below the present topographical level.

A third area of hornfelsing affects gneissose semipelites of the Tarfside Psammite Formation in a small (approximately 250 X 400 m) area immediately west of the Ladder Burn around [NO 412 844]. The hornfelses occur within 200 m of the southern contact of the Mount Battock Granite. However, in view of their high-grade nature, and since rocks immediately adjacent to the granite show little evidence for contact metamorphism, the granite would not seem to be the cause of the metamorphism in these rocks. A small aeromagnetic anomaly underlying the area of contact metamorphism (aeromagnetic map of Great Britain 1:50 000 Sheet 65E) may indicate the presence of a concealed intrusion, possibly of dioritic composition.

Conditions of thermal metamorphism

The metamorphic aureole of the Lochnagar Granite and the fringing diorite–granodiorite suite formed under a pressure of around 2.5 kbar (Goodman and Lappin, 1996). The presence of high-grade assemblages, that is prismatic sillimanite, spinel and corundum, and possible partial melting suggest that temperatures of 700 to 750°C were attained in much of the inner Glen Doll aureole, and more locally in the Crathie and Glen Gairn areas. However, the widespread development of cordierite + biotite and cordierite + K-feldspar-bearing assemblages within the hornfelses, particularly in the northern part of the Ballater district, indicates that maximum temperatures reached within the bulk of the thermal aureoles were in the range 600 ± 50°C. This estimate is based on the position of the cordierite + K-feldspar [in] reaction (Figure 19), with the upper limit being set by the quartz–muscovite breakdown reaction, which would have led to the development of andalusite + K-feldspar assemblages within the pelites. The presence of graphite can lower the temperature at which andalusite will form (Pattinson, 1989). This has led to the localised development in some andalusite pelites (for example (S77613)) of the assemblage andalusite + cordierite + muscovite.

Chapter 15 Minor intrusions

Minor intrusions, in the form of dykes, sheets and veins are a significant component of the post-tectonic igneous suite in the Ballater district. Over 300 individual intrusions have been mapped, ranging in thickness from centimetres to tens of metres and in length from metres to kilometres. They can be found throughout the district, but there are notable concentrations in the north-west and south-west segments (Figure 22).

Five distinct suites have been recognised: quartzdolerite, microdiorite, lamprophyre, felsite and microgranodiorite. This grouping differs somewhat from that of the primary survey (Barrow and Cunningham Craig, 1912) where microdiorite was listed under the heading lamprophyre etc and the microgranodiorite was not distinguished from the acid suite. The quartz-dolerite is believed to be Permo–Carboniferous in age; all others are broadly coeval with the major intrusions and hence are of late Silurian age. It would seem that microdiorites and felsites encompass rocks of more than one generation, based on the evidence of thermal metamorphism and relationships with major intrusions.

Minor intrusions within the confines of the Lochnagar Granite, which can be related directly to the various granite phases have been described in Chapter 13.

Microdiorite (PD)

Although undefined as such by Barrow and Cunningham Craig (1912), the preponderance of microdiorites in the dyke suite to the south-west of Lochnagar and around Glen Gairn was recognised by Richey (1938) and Tocher (1961) respectively. They are the most common of the intermediate intrusions in the Ballater district with over 130 known examples, the bulk of which occur north-west of a line joining Crannich Hill [NO 385 999] and Driesh [NO 271 736]. The greatest concentrations are to be found in the area north and west of Easter Balmoral, on the west side of Glen Gairn, and in Corrie Sharroch and Corrie Fee on the south side of Glen Doll (Figure 22).

Vertical sections are rare, but most of the intrusions appear to comprise near-vertical dykes. The predominant trend is north-east, although there is a range from north–south to east–west with minor north-west orientations. Most of the dykes are between 1 and 5 m in width, the largest recorded being in excess of 15 m. In most cases the dykes can be followed for only a few tens of metres, but on the west side of Glen Gairn one dyke is traceable for over 700 m.

Most dykes of the suite are microdiorites, although Tocher (1961) recorded a compositional range from pyroxene microdiorite, through microgranodiorite to biotite granophyre in the dykes of Glen Gairn. The dykes are predominantly mid-grey when fresh, the more altered examples showing a wide range of colours from pale greenish brown to purplish brown. For the most part they are medium to fine grained and porphyritic, but more equigranular to mildly porphyritic medium-grained variants occur locally.

The most common phenocrysts are plagioclase, clinopyroxene and hornblende, with less abundant biotite, quartz, sphene and magnetite. Of these, plagioclase is the dominant phase, typically forming euhedral, squat tabular to more elongate lath-like crystals which may exist as single grains or glomeroporphyritic aggregates up to 2.5 mm across. The plagioclase composition ranges from low andesine to labradorite (An36–55) with a median value of about An40. Sericitisation of plagioclase ranges from slight to extensive. In places only the outer rim is affected, indicating reaction with the groundmass.

Clinopyroxene and hornblende phenocrysts are seldom as large as plagioclase. Clinopyroxene was only recorded in a handful of rocks, being largely replaced by bundles, or more rarely single crystals, of pale green to fibrous amphibole. Primary pale brown to green hornblende is more widespread than clinopyroxene, although again much has regressed to the fibrous amphibole. Cummingtonite, with characteristic lamellar twinning, is the principal mafic phenocryst in a thin section (S92866), from a narrow dyke between the Glen Doll and Juanjorge diorites. The mineral, which occurs as aggregates of 2 to 3 grains, is also widespread in the groundmass and is mantled by biotite or, more rarely, green hornblende. The heterogeneity of the groundmass, particularly with regard to the distribution of mafic minerals, and the sporadic presence of feldspathic xenoliths suggest that the dyke may well be contaminated (compare Nockolds, 1941).

The groundmass of most dykes is an intergrowth of lath-like plagioclase with, typically, elongate brown and green hornblende, biotite and sphene. Finely divided magnetite and interstitial quartz also occur, together with accessory zircon and apatite. Sericite, epidote, chlorite and calcite are abundant in the more altered rocks. In the more acid members of the suite, K-feldspar occurs along with quartz in the intergranular phase, and more rarely (for example (S92861)) forms larger areas.

The equigranular varieties have an essentially similar petrography, comprising a medium-grained intergrowth of tabular plagioclase, rare clinopyroxene, brown or green hornblende and interstitial quartz. In many respects they are very similar to the finer grained dioritic phases of the Abergeldie Complex. For example, those rocks carrying glomeroporphyritic aggregates of hornblende (for example (S82312)) bear a striking resemblance to the 'mafic spot' rock.

Microdiorite dykes in a number of areas, including the west side of Glen Gairn, The Knock, Creag Phiobaidh, Camlet and the north side of Glen Doll, have been affected by thermal metamorphism which, except in the case of Creag Phiobaidh, can be attributed to adjacent major intrusions. The most obvious evidence is in the replacement of hornblende and clinopyroxene by ragged aggregates of foxy red-brown biotite. At more advanced stages, plagioclase loses its idiomorphic outline and recrystallises to granular aggregates with serrated intergranular boundaries. In other rocks plagioclase is 'corroded' by granular quartz. Possible cordierite was noted in one thin section (S94414). Thin microdiorite sheets, commonly dark green and highly mafic, are found in Corrie Fee and Corrie Sharrock, immediately south-west of the Glen Doll Diorite, to which they are probably related.

Lamprophyre (L, LSP, LK)

Lamprophyres are not as common as microdiorites or felsites in the Ballater district, nor are they as well exposed. They are common in the Glen Prosen–Glen Doll area but only occur sporadically in the ground east and north-east of Lochnagar. They do not appear to be spatially related to individual major plutons. The majority cut Dalradian rocks, but at [NO 2395 9969] in the gorge of the Duchrie Burn the country rock is Glen Gairn G1 granite. They usually form dykes of about 2 m in width, although within the suite there is a range from 0.2 to 10 m. Some of these intrusions are irregular in shape and terminate after a short distance. Most cannot be traced for more than 20 to 30 m, but there are a few examples of up to a few hundred metres in length, for example in Black Stripe [NO 35 87].

The majority of dykes have a north-east to north-northeast orientation, a trend broken only by three east–west dykes on the north side of Creag Mhór [NO 246 964]. Dips are mostly steep to vertical. Most lamprophyres appear to represent a single intrusive episode; however, internal chilling in the dyke from the Duchrie Burn gorge implies multiple intrusion.

The lamprophyres are mid- to dark-grey, fine- to medium-grained, equigranular to porphyritic rocks. Where they do not show the typical brick-red weathering they are difficult to distinguish from fine-grained feldspar porphyries or microdiorites. In the corries Sharroch and Fee yellow-grey-weathering, fine-grained lamprophyre dykes are present. These are mid- to dark-grey rocks which are notably tough and compact when hit with a hammer. The lamprophyres typically develop a fluted surface on weathering and usually have a thick weathered crust; they are commonly pervasively altered. In some cases, the gneissic wallrock has a narrow alteration rim, an effect not seen around the felsites.

The lamprophyres of the district are largely spessartites with phenocrysts of plagioclase, brown hornblende and ilmenite or magnetite. Green hornblende, some after clinopyroxene, is not uncommon and quartz xenocrysts are seen in some dykes. Xenocrysts, possibly derived from olivine, are present in a few rocks, but are pseudomorphed by either chlorite (S82955) or serpentine (S77692). A few of the larger dykes are kersantites with phenocrysts of biotite. The groundmass is a fine-grained intergrowth of lath-like plagioclase, hornblende, biotite and opaque minerals. Interstitial quartz and K-feldspar are present in places. The lamprophyres are locally extensively altered, and actinolite, chlorite and calcite are abundant in several specimens. The lamprophyres are moderately magnetic; kappameter readings are in the range 20 X 10−3 to 34 X 10−3 SI, although values as low as 0.3 X 10−3 SI are also recorded.

Felsite (F)

These are numerically the most abundant minor intrusions with a wide range of size and areal distribution. Like the microdiorites, they are concentrated in the neighbourhood of the major intrusions, but are also widely developed in areas remote from any exposed granitoid body, such as the south side of Glen Clova. Compared with the microdiorites they show a much greater range in form and orientation; although predominantly dyke-like and vertical, the suite includes steep-and shallow-dipping sheets. Unlike the microdiorites, the suite includes dykes which bifurcate for example at [NO 2638 9574] and [NO 2632 9682] , or have en échelon steps, as on the east-south-east ridge of Carn Dearg. Over much of the district the felsites are typically between 1 and 15 m wide, increasing to 10 to 30 m and in rare cases 50 m on the south side of Glen Clova. Being of a comparatively resistant rock the dykes commonly form distinct narrow topographical features, which may be covered extensively with angular felsite blocks. As such, some of the dykes are readily traceable for distances up to a kilometre and in one case [NO 331 991] to [NO 331 966] for nearly 2.5 km.

The orientation of these bodies seems to have been largely controlled by pre-existing structures. This is apparent not only in the sheeted bodies but also in a number of composite dykes for example at [NO 2597 9911] where the felsite has chilled against microdiorite. There is a regional variation in the strike of the felsite dykes. In the north-west part of the district, most are aligned between north-east and north-north-east; these may be representatives of the much more numerous swarm of felsite dykes which crosses the southern half of the Braemar district (Sheet 65W), and which predate or are coeval with the Lochnagar granites. In the north-east, the dominant trend is north-north-west. Around the Dee valley the north-east orientation continues to dominate, although an increasing number of dykes trend between north-east and east–west. Individual dykes on the Glen Girnock–Glen Muick watershed show a range of strike, varying over a few tens of metres from north-east to north-west. South-east of Glen Muick, the prevailing orientation of the dykes swings progressively from north–south to north-east, effectively mimicking the eastern edge of the Lochnagar Granite. This pattern is possibly the response to the stress field set up by the major intrusion. Elsewhere, dykes are seen at a high angle to the major intrusions, for example in Moulzie Burn [NO 287 787] where a number of minor felsites cut the diorite, and farther north where several porphyritic felsites have a north-west trend, almost perpendicular to the margin of the Ballater Granite [NO 37 89], with which they are probably associated. The north-east alignment persists throughout much of Glen Clova, but on the south side of the valley, where there is a preponderance of gently dipping sheets the orientation is more variable, being largely controlled by pre-existing local structures.

The felsites vary from flesh-pink to lilac to buff in colour, and from fine-grained aphanitic to distinctly porphyritic in texture. One dyke of the glassy type in Allt Chernie [NO 288 788] is flow banded, the pink and red fine-scale banding being due to slight differences in the grain size. The banding picks out highly convolute folding that occurred as the plastic melt was intruded. A number of dykes display chilled margins. A specimen (S77684) collected from the felsite portion of the composite dyke [NO 2547 9911] on the south-east side of An Creagan, displays a 14 mm-wide pink quartz–feldspar porphyry flanked by a 5 to 6 mm-wide non-porphyritic pink felsite. The latter consists almost entirely of quartz and myrmekite which is somewhat coarser grained than the matrix of the porphyry. Tocher (1961) recorded 0.3 m wide chilled margins on the 2.5 km-long dyke [NO 32 96] to [NO 33 99] mentioned above which were similarly devoid of megacrysts. From this he concluded that the megacrysts in the body of the dyke had a 'post-telluric origin'. In contrast, the 0.2 m-wide chill zone on the margin of the quartz–feldspar porphyry dyke at [NO 2585 9279] contains small phenocrysts of plagioclase and hornblende in a fine-grained purplish grey matrix.

In the majority of the porphyritic rocks the phenocrysts comprise quartz, plagioclase and K-feldspar in varying proportions. Quartz, mostly as single grains but some in aggregates of 2 to 3 grains, may be euhedral or rounded, is commonly embayed and may have rims of devitrified groundmass. Plagioclase is typically squat, tabular and euhedral; it may enclose biotite or be inter-grown with K-feldspar. Sericisation of plagioclase is common; in contrast the K-feldspar is mostly fresh, but shows a distinctive pink to red hue owing to abundant finely divided iron oxide. K-feldspar is variably perthitic, includes microcline and may be rimmed by myrmekite. Biotite, either red-brown or green, and iron oxide are less common as phenocrysts. The groundmass, which is comparable to that forming the non-porphyritic rocks, typically comprises an equigranular or, more locally, graphic intergrowth of quartz and K-feldspar with subordinate sericite and iron oxide. Plagioclase, biotite, hornblende, apatite and sphene are also present in the groundmass of some dykes.

Although dominated by acid rocks, the felsite suite also includes more intermediate examples whose composition ranges through to micromonzonite and even micromonzodiorite. They are characterised by phenocrysts of brown and green hornblende, biotite and plagioclase, and more rarely sphene and quartz. The groundmass is largely a granular aggregate of quartz, plagioclase and biotite with variable concentrations of K-feldspar and hornblende. As with the acid rocks, late quartz and K-feldspar may form graphic intergrowths. Other rocks (for example (S82320) and (S94390)) appear to represent hybrids between the acid and intermediate types, carrying phenocrysts of quartz, K-feldspar, plagioclase, hornblende and clinopyroxene. Thin section (S94390) is xenolithic, so the unusual mineralogical features could well be the result of contamination. Euhedral garnets are found within some felsites; these may be chemically resistant xenocrysts from digested country rock, or phenocrysts as seen elsewhere in Caledonian intrusions (Harrison, 1987).

Microgranodiorite (FGD)

These minor intrusions are readily distinguished from the felsite suite by their grey-white colour, speckled 'pepper and salt' appearance and general absence of phenocrysts. In places for example at [NO 2306 9630] the rock may also be foliated. The microgranodiorites consist of a medium-grained equigranular intergrowth of quartz, plagioclase (An32) and K-feldspar with subordinate biotite and accessory iron oxide and sphene. K-feldspar, including microcline, is largely interstitial to quartz and plagioclase. In a number of rocks it forms a late graphic intergrowth with quartz which may corrode earlier formed euhedral plagioclase. More rarely K-feldspar forms irregular plates which enclose quartz and plagioclase, the latter characteristically having a rim of myrmekite. Biotite forms red-brown to foxy small ragged flakes, and is partly replaced by chlorite. A few dykes have an inequigranular to weakly porphyritic texture, the coarser grained part comprising plagioclase, quartz, biotite and hornblende. The finer grained part of the rock is essentially similar, but lacks hornblende and includes sporadic K-feldspar.

Microgranodiorites are the least abundant of the minor intrusions. Most are confined to the area north and west of Crathie, reflecting their lithological affinity with the white Abergeldie Granite. The exceptions are two dykes on the col west of Tom Bad a' Mhonaidh [NO 2844 9160] and [NO 2824 9176]. The intrusions form 3 to 5 m-wide near-vertical dykes generally of limited length but locally can be followed for up to 300 m. They range in orientation from north-north-east to east-north-east. There is one exception at [NO 264 971] which is more stock-like, being 200 m long and up to 50 m wide.

Intrusion chronology

Instances of unequivocal contact relationships between minor intrusions are rare. However, in the Allt Cholzie section [NO 3492 8815] a felsite cross-cuts a lamprophyre dyke, and in the An Creagan composite dyke [NO 2541 9911] porphyritic felsite cuts and has been chilled against micro-diorite. In the White Water [NO 2336 7761] a 2 to 2.5 m-thick, grey, fine-grained lamprophyre dyke can be seen under low-water conditions to intrude a 1 to 2 m-thick aplite sheet. However, no regional time sequence can be inferred from these few examples, especially since the dyke suites may comprise intrusions of differing ages. This is most evident in the microdiorites where an earlier suite can be distinguished by the presence of hornfels textures. Such a dyke can be found on the north side of Glen Doll between [NO 257 766] and [NO 259 768]; lying within the aureole of the Glen Doll Diorite, the rock is characterised by clots of foxy red-brown biotite. All of the other examples of earlier microdiorites occur within diorite or granite for example at [NO 2309 9525] and [NO 2485 9373].

In the Glen Doll area, it is possible to attribute at least some of the minor intrusions to the major bodies. For example, those microdiorites which have not been metamorphosed by the Glen Doll Diorite relate to it; in parts they are recognisable as a fine-grained version of the diorite. Also the mineralogy of the porphyries in this part of the district reflects that of the Lochnagar L1 Granite. In the northern part of the district attribution is made more difficult by the large number of major granitoid bodies, but some broad distinctions can be made.

Many of the microdiorites and lamprophyres postdate the Lochnagar L1 and Glen Gairn G1 granites, but at only one locality [NO 2611 9225], on the western flank of Creag nan Gall, a microdiorite is seen to cut Lochnagar L3 Granite. No microdiorites were observed cutting either Lochnagar L2, Glen Gairn LGG, Ballater or Coilacriech granites. It seems probable therefore that the main episode of microdiorite–lamprophyre emplacement took place between the Lochnagar L1 and L2 granites and in view of the short time lag between the two, the dykes could well represent an early phase of the latter.

On the basis of the relationship with major intrusions, there are at least two ages of felsites and microgranites. As suggested above some are offshoots of the Lochnagar L1 Granite whereas in the centre of the complex the L2 Granite is cut by a large body of microgranite. Felsites which cut the Ballater Granite are probably an integral part of that intrusion. No such dykes are seen to cut the Coilacriech Granite.

Chapter 16 Faulting

Introduction

Faulting occurs widely in the Ballater district but is nowhere intensely developed and, with the exception of the Glen Doll and Farchal faults, none of the displacements is regionally significant. Faulting in the district has been identified by displaced strata, crush and breccia zones, quartz veining, magnetic and VLF anomalies and photolineaments. The bulk of the movements on faults seems to have occurred in the late Ordovician to early Silurian, that is after deformation of the Dalradian and before emplacement of the major intrusions, although fault reactivation has affected both major and minor intrusions. Significant fault movements appear to be largely confined to the eastern central part of the district (Figure 16). However, this conclusion is tempered by the following considerations:

Faults in the district can be divided into four groups on the basis of orientation: north-east to north-north-east, north-west, north–south and east-north-east. It is possible to establish a chronology, albeit very general, between the groups. However, it will be evident from the 1:50 000 scale map that a number of faults (most notably that which in part follows the outer edge of the Cul nan Gad Diorite) are curvilinear and as such include elements of more than one group. In broad terms, faults with north-east or north-north-east trends are the earliest, but the relationship between these and a north-west-trending set is not clear. The latter are certainly cut and displaced by faults with a north–south trend. These in turn are affected by east-north-east-trending displacements which appear to include the youngest faults in the district.

NE to NNE faults

The most extensive of these are the Glen Doll and Farchal faults. The former, can be traced for over 30 km from Glen Isla to the eastern edge of Lochnagar, although only the northernmost 12 km occurs within the Ballater district. The nature of the movement on this fault is not easy to determine, but it is likely that there has been significant dip-slip movement with downthrow to the south-east as also recorded in the Glen Shee district (Crane et al., in press). The sense and amount of strike-slip displacement is not known. The fault enters the south-east corner of the district by Cairn Dye [NO 248 722] and extends north-east along the valley of the

Burn of Kilbo. On the col at the head of this valley there is a 50 m-wide felsite dyke which appears to have been emplaced along the fault zone. A near-vertical breccia zone, 0.5 m wide, and composed of angular pieces of semipelite in a goethitic matrix, occurs at the south-east end of the col [NO 2612 7374] . This may represent a minor splay of the main fault or a parallel structure.

Approximately a kilometre to the north-east the fault line is truncated by the Glen Doll Diorite; however, locally pervasive shattering of the rock in Moulzie Burn, in the northern part of the complex, testifies to renewed movement along the line of the fault after emplacement of the diorite. North-east of the Glen Doll Diorite the fault, according to Barrow and Cunningham Craig (1912), 'traverses the metamorphic rocks for a short distance, crushing them out of all recognition'. The fault meets the outer edge of the Lochnagar L1 Granite at [NO 395 825], but there is no evidence that it continues north-eastward to the Coyles of Muick area as indicated by Barrow and Cunningham Craig (1912). Rather, the fault probably swings round to the north-west, near parallel to the structure controlling the outer edge of the Cul nan Gad Diorite, 0.4 km to the east.

The Farchal Fault lies subparallel to and roughly 2.5 km east of the Glen Doll Fault. It was not recognised during the primary survey, as there is little evidence of any disruption in the regional low-angle dip prevalent on both sides of the structure. Also in the absence of a recognised lithostratigraphical succession in this area the change in facing of the Dalradian rocks across the fault was not recognised. The main evidence for the fault's existence is that the rocks to the east are on the right-wayup median limb of the F3 Corlowie Fold, whereas those to the west are on the down faulted, upper inverted limb. An important corollary of this change in facing is that the westerly downthrow on the fault is considerably more that the 50 m vertical offset of the Longshank Gneiss Formation–metabasite boundary in the area of Sneck of the Farchal [NO 282 736] . There is additional topographical and other geological evidence for the fault. For example, the Sneck of the Farchal is a classic fault 'nick', and the southerly projection of the fault line is marked by a deeply incised valley where it cuts the southern slopes of Glen Prosen, in the Kirriemuir district to the south. On Sneck of the Farchal the Dalradian rocks are cut by a number of felsites with both intrusions and country rock showing evidence of significant movement.

Two further north-east-trending faults occur north of the Dee in the Crathie area. The first, which extends for 3.5 km north-eastwards from Craig Nordie [NO 231 948], is traceable principally as a photolineament, but locally gives rise to a breccia zone in excess of 0.5 m in width. The second forms the north-west margin of the granite on Sròn Dubh [NO 29 96] and is marked by a shallow linear depression and vein quartz float. Elsewhere, minor crush zones with this orientation are common, but there is little evidence of movement along them. North-east-trending fault breccias occur in Allt Chernie [NO 3580 9030], and to the south-west [NO 3292 8171] in a temporary trench .over the Glen Mark metadiorite. The former contains coarse pink baryte crystals, whereas the latter is characterised by epidote, garnet and minor baryte.

NW faults

The more prominent north-west-trending faults show more evidence of movement than the north-east-trending set, and are normally associated with siliceous fault breccias. For instance the fault which extends northwestwards from the Glen Doll Fault at [NO 284 778] gives rise to a 15 to 20 m-wide breccia zone in the upper reaches of the Kitchie Burn. Faults with this orientation are easily defined in the ground between the Lochnagar and Ballater granites through related topographical features (spring lines, benches etc.). The faults in this part of the district also offset magnetic units, showing clearly how the faults curve and bifurcate, with movement on one branch dying out as it is taken up elsewhere.

A swarm of north-west-trending faults, largely defined by ground magnetic surveys, is present in the area of the Water of Lee [NO 38 81]. The westernmost of these is exposed on Hunt Hill [NO 38 805], and there is good evidence in the form of gullies and blocks of vein quartz that it extends as far as [NO 372 823] the Shank of Lairs. However, it is unclear from surface evidence whether it links up with the similarly aligned fault around the Earn Stone [NO 362 838], although there is a strong linear VLF feature linking them. Also it is uncertain how it interacts with two east-north-east-trending faults that are mapped in the intervening ground. However, the VLF feature shows two distinct changes in trend at [NO 368 831] and [NO 363 837] accompanied by a diminution in amplitude. These correspond to the points of intersection with the interpreted lines of two east-north-east-trending faults (see below). There appears to be no offset of the north-west-trending fault against the east-north-east ones here, but a second north-west-trending VLF feature (possibly another fault) terminates against the trace of an east-north-east fault [NO 366 837].

A north-west-trending fault is exposed over a 30 m-long section in the Burn of Fleurs [near 389 730] on the north side of Glen Clova (Coats et al., 1993). It is marked by a 0.5 to 1.0 m-wide central clay-rich gouge containing both green and red clay alteration with shattered quartz stringers, which dips steeply to the south-west. The clay zone has a marginal alteration zone within which limonite is extensively developed, forming a saprolitic country rock in places. The structure reappears 1.5 km to the south-east in the headwaters of the Kennel Burn, where it separates schistose psammite and Cairn Trench Granite. Erosion along the line of the fault has created a 400 m-long gully averaging 10 m in depth.

A similarly aligned fault, tracing the line of the Lecht Lineament can be followed intermittently for over 10 km from the northern edge of the district [NJ 331 005]. In the north, the fault cuts the Coilacriech Granite, and to the west of Ballater controls the edge of the Ballater Granite.

N–S faults

Cutting and displacing the north-west-trending faults are several faults with north–south orientation, again more prominent in the north of the district. One such fault runs along the eastern margin of the Cul nan Gad Diorite and the sheared Glen Mark metadiorite, and is probably a late, brittle, reactivation of the ductile shear zone which deformed the metadiorite. There are several north–south faults in the east of the district, including those which form parts of the margins of the Ballater and Mount Battock granites. Movement on these faults appears to have been largely vertical and they have a considerable effect on the geology. They are marked by strong topographical features and vuggy vein quartz with hematite. The surrounding rocks are commonly altered and permeated with fine hematite, for example the red quartzites in Allt Deas [NO 3864 8764]. North–south zones of greisen alteration in the Mount Battock Granite are also related to this phase of faulting for example at [NO 397 888].

Quartz Cliff [NO 38 88] is a particularly spectacular example of a north–south fault. It is formed by a landslide whose basal slip plane follows the fault plane and the regional foliation. The fault plane is exposed for some 300 m and forms a vertical surface some 40 m in height. It contains vuggy vein quartz, with brecciated and re-cemented agates indicating several phases of movement and vein-filling. Psilomelane (hydrated manganese-barium oxide), in addition to hematite, is commonly associated with the quartz. The fault can be traced farther to the south through quartz boulders on the hillside, but to the north the fault must die out, as it is not seen in the well-exposed quartzites in Allt Deas.

ENE faults

Some north–south faults are displaced by east-north-easttrending displacements which represent the last main phase through much of the district, with some of the major structures running for several kilometres. These faults have a small amount of lateral throw, commonly showing a sinistral sense of displacement. They are most evident where they displace magnetic boundaries, but are not marked by breccias or topographical features. They are also readily detectable by VLF–EM surveys. For example, this method confirmed the east-north-east-trending fault cutting the magnetic Glen Mark metadiorite to the north of Black Hill of Mark [NO 327 822], and demonstrated the continuity of it, and two other parallel faults across poorly exposed ground farther to the east. On (Figure 16), the east-north-east faults appear to be more abundant in the south, although this is probably due to the fact that a detailed magnetic survey was conducted there, and that there are a number of magnetic marker horizons present.

Chapter 17 Devonian

Pisolitic silcrete

On the watershed between Boustie Ley [NO 322 760] and Benty Roads [NO 330 766] the Longshank Gneiss Formation is reddened and contains a great deal of fine hematite and jasper, and float within wasted peat locally includes blocks and fragments of a pisolitic silcrete. The silcrete is not exposed, but its outcrop is interpreted from the available evidence. It probably represents a Devonian land surface which here is locally coincident with the present topography.

The silcrete occurs as stratified blocks up to 30 cm across, though seldom more than 15 cm thick. A rough bedding can be seen in the blocks, with beds a few centimetres thick showing differences in clast type and abundance, different grain size and varying amounts of reddening. The matrix is commonly a fine red chert or jasper, and a coating of this on the grains gives the rock a red colour on weathered surfaces. The grains comprise both rounded pisoliths and more angular fragments, the latter usually being the larger. The pisoliths range from 1 to 7 mm in diameter, the sphericity of the grains decreasing with increasing grain size. Many of the smaller pisoliths have a concentric internal structure, generally visible to the naked eye on a cut surface, and are silicic versions of caliche type concretions. Larger fragments may contain quartz and feldspar, possibly with some muscovite present; these are clasts of the local gneiss. Fine white rims are present around some of the clasts, and in thin section these are seen to be fringes of chert needles radiating from their margins. Some flakes of muscovite derived from the gneiss occur within the chert, the original muscovite grains having been divided by the formation of fine chert needles between the mica layers.

The Longshank Gneiss Formation is (Chapter 8) characterised by abundant magnetite. In the area where the silcrete float is found the magnetite is altered to hematite which locally forms dendritic masses. The hematisation causes a significant drop in the magnetic susceptibility from the generally high values (1 to 10 SI). The altered rock no longer has an interlocking texture, but between grains and aggregates of grains chert needles radiate out from the grain margins. In places 'islands' of gneiss have formed, completely surrounded by chert. The biotite has been destroyed, but muscovite grains are present in which chert needles separate the individual flakes. In this way the fresh gneiss is broken down to form the basis of fragments in the silcrete.

A felsite dyke on Benty Roads [NO 332 764] has been similarly altered. The dyke is dark red in colour due to the presence of disseminated hematite, which has evidently developed from the original magnetite. The dyke is some 500 m in length, and even where it intrudes gneisses which are not particularly reddened, it has been extensively altered, suggesting the dyke reacted more readily than the surrounding gneiss, perhaps owing to circulation paths of oxidising fluids.

The presence of clasts of local Dalradian gneiss, of hematite in the chert cement and of the concentrically zoned pisoliths are indicative of the formation of a caliche-type soil on the Dalradian land surface, at a time when semi-arid conditions prevailed and the rate of evaporation exceeded that of precipitation. The deposit here is highly siliceous, and could either have formed as a silcrete, or as a calcrete, later silicified. It seems likely that deep weathering of the Dalradian gneisses released silica into solution, and oxidised magnetite in the bedrock to hematite. As excessive evaporation occurred, the silica was deposited in the soil profile and in weathered bedrock as the concentric shells of pisoliths, chert needles around rock fragments and within mica grains, and as amorphous chalcedony between the clasts.

The deposition of the silcrete horizon during a period of subaerial exposure obviously occurred later than the last of the Caledonian deformations, and the alteration of the felsite dyke indicates that it was no earlier than Silurian–Devonian times (Rock et al., 1988). Silcrete horizons are found in sedimentary rocks of both Devonian and Triassic age in Scotland. The sparsity of Triassic deposits on the Scottish mainland compared with Old Red Sandstone suggests that this silcrete is more likely to be of Devonian age. The Devonian outcrops closest to Benty Roads are those in the Midland Valley and in fault-bounded basins directly to the north of the Highland Boundary Fault (Cameron and Stephenson, 1985). There are also several outliers of Old Red Sandstone sedimentary rocks in north-east Scotland, for example at Rhynie and Tomintoul (Mykura, 1983). The situation of these outliers as preserved areas of Devonian strata lying unconformably on Dalradian basement, has more in common with that of the Benty Roads silcrete than has the geographically closer fault-bounded basin of the Midland Valley. Silcretes are present in the Old Red Sandstone of Scotland, mainly within the lacustrine sedimentary rocks of the Orcadian basin (Parnell, 1983). Some of these are the result of algal activity, but pedogenic silcretes with textures typical of soil profiles, comparable to the Benty Roads example, are also found.

The presence of the silcrete above the glacial trough of Glen Clova indicates that, during the Devonian, in the uplifted area between the Midland Valley and Orcadian basins there were occasionally areas with subaerial weathering and silcrete formation, rather than erosion and removal of the basement rock. Thus by the time the silcrete was formed uplift of the Grampians had virtually ceased, and the Caledonian land surface had been eroded approximately to the level seen at the present day.

Information sources

Geological information held by the British Geological Survey relevant to the Ballater district is listed below. It includes published material in the form of maps, memoirs and reports and unpublished maps and reports. Also included are other sources of data held by BGS in a number of collections, including borehole records, mine plans, fossils, rock samples, thin sections, hydrogeological data and photographs.

Published maps and reports are listed in the BGS Catalogue of geological maps and books and the Geoscience Data Index can be accessed at: http://www.bgs.ac.uk

BGS Maps

1:10 000 and 1:10 560

The maps at 1:10 000 scale covering wholly or in part in the 1:50 000 scale Sheet 65E are listed below, together with the surveyors' initials and the dates of the survey. The surveyors were: M J Gallagher, S Goodman (University of Aberdeen), D Gould, A G Leslie (Queen's University, Belfast), D I J Mallick, W J McCourt, J R Mendum, S Robertson and C G Smith. All maps are Solid only. The maps are not published but are available for consultation in the BGS Library, Edinburgh, and also at Keyworth and the London Information Office, in the Natural History Museum, South Kensington, London. Dyeline copies may be purchased from the Sales Desk.

Sheet Surveyor Published
NJ 30 SW MJG, SR 1991
NO 27 NW JRM 1991
NO 27 NE JRM, SR, CGS, SG 1991
NO 27 SW SR 1994
NO 27 SE JRM, SR, CGS 1992
NO 28 SE SG 1989
NO 29 NW WJM, CGS 1992
NO 29 NE SR, CGS 1992
NO 29 SW DIJM, WJM, CGS 1992
NO 29 SE DIJM, CGS 1992
NO 37 NW SR, CGS, SG 1991
NO 37 NE SR 1991
NO 37 SW SR, CGS 1992
NO 37 SE SR 1991
NO 38 NW SG, DIJM, CGS 1992
NO 38 NE SG 1989
NO 38 SE SG, SR 1991
NO 39 NW SR, CGS 1992
NO 39 NE SR, CGS 1992
NO 39 SW SG, CGS 1992
NO 39 SE SG, CGS 1989
NO 47 SW DG, SR 1992
NO 48 SW DG, SR 1992
NO 48 NW SG 1989
NO 49 NW DG, SR, CGS 1991
NO 49 SW SG 1989

BGS books

BGS Registered Number National Grid Vertical Depth
(NO49NW/1) [NO 4010 986] 298.25 m

Site exploration reports

This collection consists of site exploration reports carried out to investigate foundation conditions prior to construction. There is a digital index and the reports themselves are held on microfiche.

Hydrogeological data

Records of water boreholes are held at BGS, Edinburgh.

Geochemical data

Records of stream-sediment and other analyses are held at BGS, Edinburgh.

Gravity and magnetic data

Records are held at BGS, Edinburgh.

Seismic data

Records of earthquakes are held at BGS, Edinburgh.

Material collections

Geological Survey photographs

Some 80 photographs illustrating aspects of the geology of the Ballater district are deposited for reference in the libraries at BGS, Edinburgh, and BGS, Keyworth; and in the BGS Information Office, London. These are: D3770–3785, 4252–4268, 5253–5258, 5260–5292, 5345/11, 5345/13–16, D5346/24, D5346/26–32. Most of the photographs were taken over the last two decades. The photographs depict details of the various rocks and sediments exposed, either naturally or in excavations and also some general views. The photographs can be supplied as black and white or colour prints and 2 X 2 colour transparencies, at a fixed tariff, from the Photographic Department, BGS, Edinburgh.

Petrological collections

The petrological collections for the Ballater district consist of about 2000 hand specimens and thin sections, prepared from rocks collected by BGS and university personnel working on the survey of the district and made at Aberdeen University as part of the mapping contract; all are held at BGS Edinburgh. The samples and thin sections are of the igneous and metamorphic rocks in the district. Information on databases of rock samples, thin sections and geochemical analyses can be obtained from the Group Manager, Mineralogy and Petrology Section, BGS, Edinburgh.

Other relevant data

Sites of Special Scientific Interest are the responsibility of Scottish Natural Heritage, Battleby Redgorton, Perth, PH1 3EW.

Remote sensing data Aerial photography for Scotland is held at the Royal Commission on the Ancient and Historical Monuments of Scotland, Edinburgh. Vertical stereo pairs for the district, at a scale of approximately 1:23 000, flown in 1968–69, are available for consultation there. More recent photography on a similar scale is only available commercially.

References

Most of the references listed below are held in the Library of the British Geological Survey at Murchison House, Edinburgh and at Keyworth, Nottingham. Copies of the references can be purchased subject to the current copyright legislation.

ANDERSON, J G. 1942. The stratigraphical order of the Dalradian schists near the Highland Border in Angus and Kincardine. Transactions of the Geological Society of Glasgow, Vol. 20, 223–237.

ANDERTON, R. 1985. Sedimentation and tectonics in the Scottish Dalradian. Scottish Journal of Geology, Vol. 21, 407–436.

ASHCROFT, W A. 1958. The rocks between Red Craig and Loch Brandy, Glen Clova. Unpublished BSc thesis, University of Aberdeen.

ASHCROFT, W A, KNELLER, B C, LESLIE, A G, and MUNRO, M. 1984. Major shear zones and autochthonous Dalradian in the northeast Scottish Caledonides. Nature, London, Vol. 310, 760–762.

BAKER, A J. 1985. Pressures and temperatures of metamorphism in the eastern Dalradian. Journal of the Geological Society of London, Vol. 142, 137–148.

BAKER, A J, and DROOP, G T R. 1983. Grampian metamorphic conditions deduced from mafic granulites and sillimanite K-feldspar gneisses in the Dalradian of Glen Muick, Scotland. Journal of the Geological Society of London, Vol. 140, 487–497.

BARROW, G. 1893. On an intrusion of muscovite-biotite gneiss in the southeast Highlands of Scotland and its accompanying metamorphism. Quarterly Journal of the Geological Society of London, Vol. 19, 33–58.

BARROW, G. 1912. On the geology of lower Deeside and the southern Highland border. Proceedings of the Geologists' Association, Vol. 23, 274–290.

BARROW, G, and CUNNINGHAM CRAIG, E H. 1912. The geology of the districts of Braemar, Ballater and Glen Clova. (Explanation of Sheet 65.) Memoir Geological Survey of Scotland. (HMSO.)

BOUTCHER, S M A. 1963. The geology of the Geallaig district, Upper Deeside. Unpublished PhD thesis, University of Edinburgh.

BRITISH GEOLOGICAL SURVEY. 1991. Regional geochemistry of the East Grampians area. (Keyworth, Nottingham: British Geological Survey.)

CAMERON, I B, and STEPHENSON, D. 1985. British Regional Geology: The Midland Valley of Scotland. (3rd edition). (London: HMSO for British Geological Survey.)

CHACKSHELD, B C, SHAW, M H, COATS, J S, SMITH C G, and STEPHENSON, D. 1997. Exploration for stratabound mineralisation in the Argyll Group (Dalradian) of north-east Scotland. Mineral Reconnaissance Program Report, British Geological Survey, No. 145.

CHINNER, G A. 1960. Pelitic gneisses with varying ferrous/ferric ratios from Glen Clova, Angus, Scotland. Journal of Petrology, Vol. 1, 178–217.

CHINNER, G A. 1961. The origin of sillimanite in Glen Clova, Angus. Journal of Petrology, Vol. 2, 312–323.

CHINNER, G A. 1965. The kyanite isograd in Glen Clova, Angus, Scotland. Mineralogical Magazine, Vol. 34, 132–143.

CHINNER, GA. 1966. The distribution of pressure and temperature during Dalradian metamorphism. Quarterly Journal of the Geological Society of London, Vol. 122, 159–186.

CHINNER, G A. 1980. Kyanite isograds of Grampian metamorphism. Journal of the Geological Society of London, Vol. 137, 35–39.

CHINNER, G A, and HESELTINE, Pir. 1979. The Grampide andalusite/kyanite isograd. Scottish Journal of Geology, Vol. 15, 117–127.

COATS, J S, FORTEY, N J, GALLAGHER, M J, and GROUT, A. 1984. Stratiform barium enrichment in the Dalradian of Scotland. Economic Geology, Vol. 79, 1585–1595.

COATS, J S, SHAW, M H, ROLLIN, K E, ROBERTSON, S, and REDWOOD, S D. 1993. Mineral exploration in the Pitlochry to Glen Clova area, Tayside Region, Scotland. Mineral Reconnaissance Programme Report, British Geological Survey, No. 126.

COBBING, E J, PITFIELD, P J, DERBYSHIRE, D P F, and MALLICK, D I J. 1992. The granites of the south-east Asian tin belt. Overseas Memoir of the British Geological Survey, No. 10.

COX, F C. 1966. Structure and metamorphism of the area between Clunie Lodge and Cairnwell, Aberdeenshire. Unpublished PhD thesis, University of London, Bedford College.

CRANE, A. 1987. The geology of Deeside. 227–242 in Excursion guide to the geology of the Aberdeen area. TREWIN, N H, KNELLER, B C, and GILLEN, C (editors). (Edinburgh: Scottish Academic Press for Geological Society of Aberdeen.)

CRANE, A, GOODMAN, S, KRABBENDAM, M, LESLIE, A G, ROBERTSON, S, and ROLLIN, K. In Press. Geology of the Glen Shee district, Scotland. (Sheet 56W and adjacent areas). Memoir of the British Geological Survey.

DEMPSTER, T J. 1985. Uplift patterns and orogenic evolution in the Scottish Dalradian. Journal of the Geological Society of London, Vol. 142, 111–128.

DUNHAM, K C. 1952. Fluorspar (4th edition). Special Reports on the Mineral Resources of Great Britain. Memoir of the Geological Survey of Great Britain.

FERRY, J M, and SPEAR, F S. 1978. Experimental calibration of the partitioning of Fe and Mg between biotite and garnet. Contributions to Mineralogy and Petrology, Vol. 66, 113–117.

FETTES, D J, GRAHAM, C M, HARTE, B, and PLANT, J A. 1986. Lineaments and basement domains; an alternative view of Dalradian evolution. Journal of the Geology Society of London, Vol. 143, 453–464.

GAMMON, J B. 1966. Fahlbands in the Precambrian of southern Norway. Economic Geology, Vol. 61, 174–188.

GOODMAN, S. 1991. The Pannanich Hill Complex and the origin of Crinan Subgroup migmatites in the North-eastern and Central Highlands. Scottish Journal of Geology, Vol. 27, 147–156.

GOODMAN, S. 1993. Survival of zincian staurolite to upper amphibolite facies metamorphic grade. Mineralogical Magazine, Vol. 57, 736–739.

GOODMAN, S. 1994. The Portsoy-Duchray Hill Lineament; a review of the evidence. Geological Magazine, Vol. 130, 407–415.

GOODMAN, S, and LAPPIN, M A. 1996. The thermal aureole of the Lochnagar Complex: mineral reactions and implications for thermal modelling. Scottish Journal of Geology, Vol. 32, 159–172.

GOODMAN, S, LESLIE, A G, ASHCROFT, W A, and CRANE, A. 1990. The geology of the central part of Sheet 65E (Ballater); contribution to the memoir. British Geological Survey Technical Report, WA/90/59.

GOODMAN, S, and WINCHESTER, J A. 1993. Geochemical variations within metavolcanic rocks of the Dalradian Farragon Beds and adjacent formations. Scottish Journal of Geology, Vol. 29, 131–141.

GOULD, D. 1997. Geology of the country around Inverurie and Alford. Memoir of the British Geological Survey, Sheets 76E and 76W (Scotland).

GOULD, D. 2001. Geology of the Aboyne district. Memoir of the British Geological Survey, Sheet 66W (Scotland).

GRAHAM, C M. 1976a. Petrochemistry and tectonic significance of Dalradian metabasaltic rocks of the SW Scottish Highlands. Journal of the Geological Society of London, Vol. 132, 61–84.

GRAHAM, C M. 1976b. Petrochemical affinities of Dalradian metabasaltic rocks: discussion of a paper by J A Winchester and P A Floyd. Earth and Planetary Science Letters, Vol. 32, 210–212.

GUNN, A G, STYLES, M T, ROLLIN, K E, and STEPHENSON, D. 1996. The geology of the Succoth-Brown Hill mafic-ultramafic intrusive comlex, near Huntly, Aberdeenshire Scottish Journal of Geology, Vol. 32, 33–50.

HALL, A, and WALSH, J N. 1972. Zinnwaldite granite from Glen Gairn, Aberdeenshire. Scottish Journal of Geology, Vol. 8, 265–267.

HALLIDAY, A N, AFTALION, M, VAN BREEMEN, O, and JOCELYN, J. 1979. Petrogenetic significance of Rb-Sr and U-Pb isotopic systems in the 400 Ma old British Isles granitoids and their hosts. 653–662 in The Caledonides of the British Isles reviewed. HARRIS, A L, HOLLAND, C H, and LEAKE, B E (editors). Special Publication of the Geological Society of London, No. 8.

HARRIS, A L, BRADBURY, H J, and MCGONIGAL, N H. 1976. The evolution and transport of the Tay Nappe. Scottish Journal of Geology, Vol. 12, 103–113.

HARRIS, A L, BALDWIN, C T, BRADBURY, H J, JOHNSON, H D, and SMITH, R A. 1978. Ensialic basin sedimentation: the Dalradian Supergroup. 115–138 in Special Issue of the Geological Journal No. 10 Crustal evolution in northwestern Britain. BOWES, D R, and LEAKE, B E (editors). (Liverpool: Seel House Press.)

HARRIS, A L, HASELOCK, P J, KENNEDY, M J, and MENDUM, J R. 1994. The Dalradian Supergroup in Scotland, Shetland and Ireland. 33–53 in A revised correlation of Precambrian rocks in the British Isles. GIBBONS, W, and HARRIS, A L (editors). Special Report of the Geological Society of London, No. 22.

HARRIS, A L, and PITCHER, W S. 1975. The Dalradian Supergroup. 52–75 in A correlation of Precambrian Rocks of the British Isles. HARRIS, A L, and others (editors). Special Report of the Geological Society of London, No. 6.

HARRISON, T N. 1987. The Eastern Grampians Granites. Unpublished PhD thesis, University of Aberdeen.

HARRISON, T N, and HUTCHINSON, J. 1987. The age and origin of the Eastern Grampians Newer Granites. Scottish Journal of Geology, Vol. 23, 269–282.

HARRY, W T. 1958. A re-examination of Barrow's Older Granites in Glen Clova. Transactions of the Royal Society of Edinburgh, Vol. 53, 393–412.

HARTE, B. 1966. Stratigraphy, structure and metamorphism in the south-eastern Grampian Highlands of Scotland. Unpublished PhD thesis, University of Cambridge.

HARTE, B. 1979. The Tarfside succession and the structure and stratigraphy of the eastern Scottish Dalradian rocks. 221–228 in The Caledonides of the British Isles - reviewed. HARMS, A L, HOLLAND, C H, and LEAKE, B E (editors). Special Publication of the Geological Society of London, No. 8.

HARTE, B, BOOTH, J E, DEMSTER, T J, FETTES, D J, MENDUM, J R, and WATTS, D. 1984. Aspects of the post-depositional evolution of Dalradian and Highland Border Complex rocks in the Southern Highlands of Scotland. Transactions of the Royal Society of Edinburgh: Earth Sciences, Vol. 75, 151–163.

HARTE, B, and DEMPSTER, T J. 1987. Regional metamorphic zones: tectonic controls. Philosophical Transactions of the Royal Society of London, Vol. 321, 105–27.

HARTE, B, and HUDSON, N F C. 1979. Pelite facies series and the temperatures and pressures of Dalradian metamorphism. 323–337 in The Caledonides of the British Isles - reviewed. HARRIS, A L, HOLLAND, C H, and LEAKE, B E (editors). Special Publication of the Geological Society of London, No. 8.

HARTE, B, and JOHNSON, M R W. 1969. Metamorphic history of Dalradian rocks in Glens Clova, Esk and Lethnot, Angus, Scotland. Scottish Journal of Geology, Vol. 5, 54–80.

HAWSON, C A, and HALL, A J. 1987. Middle Dalradian Corrycharmaig serpentinite, Perthshire, Scotland: an ultramafic intrusion. Transactions of the Institution of Mining and Metallurgy (Section B: Applied Earth Science), Vol. 96, B173–177.

HEDDLE, M F. 1877. Chapters on the mineralogy of Scotland. Transactions of the Royal Society of Edinburgh, Vol. 28, 249–299.

HEDDLE, M F. 1901. The mineralogy of Scotland. (Edinburgh: David Douglas).

HODGES, KV, and SPEAR, F S. 1982. Geothermometry, geobarometry and the A L2 Si05 triple point at Mt Moosilauke, New Hampshire. American Mineralogist, Vol. 67, 118–1134.

HOLDAWAY, M J. 1971. Stability of andalusite and the aluminium-silicate phase diagram. American Journal of Science, Vol. 271, 97–131.

HUDSON, N F C. 1980. Regional metamorphism of some Dalradian pelites in the Buchan area. Contributions to Mineralogy and Petrology, Vol. 73, 39–51.

HUDSON, N F C. 1985. Conditions of Dalradian metamorphism in the Buchan area, NE Scotland. Journal of the Geological Society of London, Vol. 142, 63–76.

HUTCHISON, A G. 1933. The metamorphism of the Deeside Limestone, Aberdeenshire. Transactions of the Royal Society of Edinburgh, Vol. 57, 557–592.

JACQUES, J M, and REAVY, R J. 1994. Caledonian plutonism and major lineaments in the SW Scottish Highlands. Journal of the Geological Society of London, Vol. 151, 955–969.

JARVIS, K E. 1987. The petrogenesis and geochemistry of the diorite complexes south of Balmoral Forest, Angus. Unpublished PhD thesis, City of London Polytechnic.

JOHNSTONE, G S. 1966. British Regional Geology: The Grampian Highlands. (3rd edition). (Edinburgh: HMSO for Geological Survey and Museum.)

KNELLER, B C, and LESLIE, A G. 1984. Amphibolite facies metamorphism in shear zones in the Buchan area of NE Scotland. Journal of Metamorphic Geology, Vol. 2, 83–94.

KNELLER, B C, and AFTALION, M. 1987. The isotopic and structural age of the Aberdeen Granite. Journal of the Geological Society of London, Vol. 144, 717–721.

LAWRIE, T R M. 1943. Report on the old lead mine at Abergairn, near Balloter, Aberdeenshire and report on the old lead workings at Monaltrie, near Ballater, Aberdeenshire, (unpublished). Mineral Resource Files, British Geological Survey, Edinburgh.

LEE, M K. 1984. Analysis of geophysical logs from the Shap, Cairngorm, Ballater, Mount Battock and Bennachie boreholes. Investigation of the geothermal potential UK British Geological Survey Geothermal Energy Research Programme: Energy Resources, WJ/FE/84/23.

LISTER, G S, and SNOKE, A W. 1984. S-C mylonites. Journal of Structural Geology, Vol. 6, 617–638.

MACGREGOR, A G. 1929. Metamorphism around the Lochnagar Granite, Aberdeenshire. Report of the British Association for 1928, 553.

MACGREGOR, A G. 1948. British Regional Geology: the Grampian Highlands (2nd edition). (Edinburgh: HMSO for Geological Survey and Museum.)

MAHMOUD, L A. 1986. Mineralogy, petrology and geochemistry of some zoned dioritic complexes in Scotland. Unpublished PhD thesis, University of St Andrews.

MCLELLAN, E L. 1982. Barrovian migmatites and the thermal history of the south-eastern Grampians. Unpublished PhD thesis, University of Cambridge.

MCLELLAN, E L. 1985. Metamorphic reactions in the kyanite and sillimanite zones of the Barrovian type area. Journal of Petrology, Vol. 26, 789–818.

MCLELLAN, E L. 1989. Sequential formation of subsolidus and anatectic migmatites in response to thermal evolution, eastern Scotland. Journal of Geology, Vol. 97, 165–82.

MENDUM, J R, and FETTES, D J. 1985. The Tay nappe and associated folding in the Ben Ledi-Loch Lomond area. Scottish Journal of Geology, Vol. 21, 41–56.

MYKURA, W. 1983. Old Red Sandstone. 205–252 in Geology of Scotland (2nd edition). CRAIG, G Y (editor). Edinburgh: Scottish Academic Press.

NEWTON, R C, and HASELTON, H T. 1981. Thermodynamics of the garnet-plagioclase-A L2 Si05-quartz geobarometer. 131–147 in Thermodynamics of mineral and melts. NEWTON, R C, NAVROTSKY, A, and WOOD, B J (editors). Advances in Physical Geochemistry, Vol. 1. (New York: Springer-Verlag.)

NOCKOLDS, S R. 1941. The Garabal Hill-Glen Fyne igneous complex. Quarterly Journal of the Geological Society of London, Vol. 96, 451–511.

O'BRIEN, C. 1985. The petrogenesis and geochemistry of the British Caledonian granites, with special reference to mineralized intrusions. Unpublished PhD thesis, University of Leicester.

OLDERSHAW, W A. 1958. A sudy of the Lochnagar granitic ring complex, Aberdeenshire. Unpublished PhD thesis, University of London, Queen Mary College.

OLDERSHAW, W. 1974. The Lochnagar granitic ring complex, Aberdeenshire. Scottish Journal of Geology, Vol. 10, 297–309.

ORREDGE, G R. 1961. A study of the granitic and associated rocks of the Glen Gairn complex, Aberdeenshire. Unpublished PhD thesis, University of London, Queen Mary College.

PANKHURST, R J. 1974. Rb-Sr whole-rock chronology of Caledonian events in northeast Scotland. Bulletin of the Geological Society of America, Vol. 85, 345–350.

PARNELL, J. 1983. Ancient duricrusts and related rocks in perspective; a contribution from the Old Red Sandstone. 179–209 in Residual deposits: surface related weathering processes and materials. WILSON, R C L (editor). Special Publication of the Geological Society of London, No. 11.

PATTINSON, D R M. 1989. P-T conditions and the influence of graphite on pelitic phase relations in the Ballachulish aureole, Scotland. Journal of Petrology, Vol. 30, 1219–1244.

PEACOCK, J D, CLARK, G C, MAY, F, MENDUM, J R, ROSS, D L, and RUCKLEY, A E. 1977. Sand and gravel resources of Grampian Region. Report of the Institute of Geological Science, Vol. 77/2.

PEARCE, J A, HAINES, N B W, and TINDLE, A G. 1984. Trace element discrimination diagrams for the tectonic interpretation of granitic rocks. Journal of Petrology, Vol. 25, 956–983.

PHILLIPS, E R. 1991. Petrology of the Lochnagar and Glen Gairn igneous complexes and associated diorite-granodiorite suites, Sheet 65E (Ballater), Scotland. British Geological Survey Technical Report, WG/91/14.

PHILLIPS, E R. 1992a. Mineralogy and petrology of a series of thermal metamorphosed pelites and associated rock types, Sheet 65E (Balloter), Scotland. British Geological Survey Technical Report, WG/92/37.

PHILLIPS, E R 1992b. Mineralogy and petrology of the Glen Doll and Glen Mark thermal aureoles, East Grampian Highlands. British Geological Survey Technical Report, WG/92/17.

PIASECKI, M A J, and VAN BREMMEN, O. 1983. Field and isotopic evidence for a c. 750 Ma tectonothermal event in Moine rocks in the Central Highland region of the Scottish Caledonides. Transactions of the Royal Society of Edinburgh: Earth Sciences, Vol. 73, 119–134.

PIDGEON, R T, and AFTALION, M. 1978. Cogenetic and inherited zircon U-Pb systems in granites: Palaeozoic granites of Scotland and England. 183–220 in Crustal Evolution in northwestern Britain and adjacent areas. BOWES, D R, and LEAKE, B E (editors). (Glasgow: Scottish Academic Press.)

PLANT, J A, HENNEY, P J, and SIMPSON, P R. 1990. The genesis of tin-uranium granites in the Scottish Caledonides: implications for metallogenesis. Geological Journal, Vol. 25, 431–442.

READ, H H. 1919. The two magmas of Strathbogie and Lower Banffshire. Geological Magazine, Vol. 56, 364–371.

READ, H H. 1928. The Highland Schists of middle Deeside and east Glen Muick. Transactions of the Royal Society of Edinburgh, Vol. 55, 755–772.

READ, H H. 1952. Metamorphism and migmatisation in the Ythan Valley, Aberdeenshire. Transactions of the Edinburgh Geological Society, Vol. 15, 265–279.

READ, H H. 1961. Aspects of the Caledonian magmatism in Britain. Proceedings of the Liverpool and Manchester Geological Society, Vol. 2, 653–683.

RENNIE, F W. 1983. The mineralogy and geochemistry of the Lochnagar complex. Unpublished PhD thesis, University of Aberdeen.

RICHARDSON, S W, GILBERT, M C and BELL, P M. 1969. Experimental determination of kyanite-andalusite and andalusite-silliminanite equilibria; the aluminium silicate triple point. American Journal of Science, Vol. 267, 259–272.

RICHEY, J E. 1938. The dykes of Scotland. Transactions of the Edinburgh Geological Society, Vol. 13, 393–435.

RIDDLER, G. 1967. The geology of an area to the SW of Lochnagar, West Aberdeenshire. Unpublished BSc thesis, University of Aberdeen.

ROBERTS, J L. 1966. Sedimentary affiliations and stratigraphic correlation of the Dalradian rocks in the south-west Highlands of Scotland. Scottish Journal of Geology, Vol. 2, 200–223.

ROBERTSON, S. 1988. Relationships between 'younger' and 'older basics' in the Glen Gairn area, North Deeside. Scottish Journal of Geology, Vol. 24, 89–92.

ROBERTSON, S. 1991. Older granites in the south-eastern Scottish Highlands. Scottish Journal of Geology, Vol. 27, 21–26.

ROBERTSON, S. 1994. Timing of Barrovian metamorphism and 'Older Granite' emplacement in relation to Dalradian deformation. Journal of the Geological Society of London, Vol. 151, 5–8.

ROBERTSON, S, and SWAINBANK, I. 1990. The age and isotopic characteristics of the older granites in the Glen Clova-Glen Esk area, south-east Scottish Highlands. British Geological Survey Technical Report, WA/90/82.

ROGERS, G, DEMPSTER, T J, BLUCK, B J, and TANNER, P W G. 1989. A high-precision U/Pb age for the Ben Vuirich granite: implications for the evolution of the Scottish Dalradian Supergroup. Journal of the Geological Society of London, Vol. 146, 789–798.

ROGERS, G, PATERSON, B A, DEMPSTER, T J, and REDWOOD, S D. 1994. U-Pb geochronology of the 'Newer' gabbros, NE Grampians. In Caledonian terrane relationships in Britain. (Keyworth, Nottingham: Symposium Abstracts, British Geological Survey.)

ROLLIN, K E. 1984. Gravity modelling of the Eastern Highlands granites in relation to heat flow studies. Investigation of the geothermal potential of the UK. British Geological Survey Geothermal Energy Research Programme: Energy Resources, WJ/GE/84/15.

RUSSELL, A. 1937. Notes on the occurrence of fluorite in Aberdeenshire and Banffshire. Mineralogical Magazine, Vol. 24, 307.

SCHUMACHER, R. 1985. Zincian staurolite in Glen Doll. Mineralogial Magazine, Vol. 49, 561–571.

SMITH, C G, GALLAGHER, M J, COATS, J S, and PARKER, M E. 1984. Detection and general characteristics of stratabound mineralization in the Dalradian of Scotland. Transactions of the Institution of Mining and Metallurgy (Section B: Applied Earth Science), Vol. 93, B125–133.

SOPER, N J, and ENGLAND, R W.1995. Vendian and Riphean rifting in NW Scotland. Journal of the Geological Society of London, Vol. 152, 11–14.

STARKEY, R E. 1990. Note: Bentradite from the Pass of Ballater, Grampian Region, Scotland. Journal of the Russell Society, Vol. 3, 31.

STEPHENS, W E, and HALLIDAY, A N. 1984. Geochemical contrasts between late Caledonian granitoid plutons of northern, central and southern Scotland. Transactions of the Royal Society of Edinburgh, Earth Sciences, Vol. 75, 259–273.

STEPHENSON, D, and GOULD, D. 1995. British Regional Geology: The Grampian Highlands. (4th edition). (London: HMSO for the British Geological Survey.)

TADESSE, T. 1991. A stratigraphic, structural and metamorphic analysis of Dalradian rocks west of Ballater, NE Scotland. Unpublished PhD thesis, Queen's University of Belfast.

TANNER, P W G, and LESLIE, A G. 1994. A pre- D2 age for the 590 Ma Ben Vuirich Granite from the Dalradian of Scotland. Journal of the Geological Society of London, Vol. 151, 209–212.

TINDLE, A G, and WEBB, P C. 1989. Niobian wolframite from Glen Gairn, Eastern Highlands of Scotland. Geochimica et Cosmochimica Acta., Vol. 53, 1921–1935.

TINDLE, A G, and WEBB, P C. 1990. Estimation of lithium contents in trioctahedral micas using microprobe data: application to micas from granitic rocks. European Journal of Mineralogy, Vol. 2, 595–610.

TOCHER, F E. 1958. The igneous and metamorphic rocks of the eastern end of the Geallaig Hill, West Aberdeenshire. Unpublished PhD thesis, University of Aberdeen.

TOCHER, F E. 1961. Microdiorite and other dykes of lower

Glen Gairn, near Ballater, west Aberdeenshire. Transactions of the Edinburgh Geological Society, Vol. 18(3), 230–239.

UPTON, P S. 1986. A structural cross-section of the Moine and Dalradian rocks of the Braemar area. Report of the British Geological Survey, Vol. 17, No. 1, 9–19.

VAN BREEMEN, O, and BOYD, R. 1972. A radiometric age for pegmatite cutting the Belhelvie mafic intrusion, Aberdeenshire. Scottish Journal of Geology, Vol. 8, 115–120.

VAN DE KAMP, P C. 1970. The Green Beds of the Scottish Dalradian Series: geochemistry, origin and metamorphism of mafic sediments. Journal of Geology, Vol. 78, 281–303.

WEBB, P C, and BROWN, G C. 1984. The Eastern Highlands granites: heat production and related geochemistry. Investigations of the geothermal potential of the UK. British Geological Survey Geothermal Energy Research Programme: Energy Resources, WJ/GE/84/13.

WEBB, P C, TINDLE, A G, and IXER, R A. 1992. W-Sn-Mo-Bi-Ag mineralization associated with zinnwaldite granite from Glen Gairn, Scotland. Transactions of the Institution of Mining and Metallurgy (Series B: Applied Earth Sciences), Vol. 101, 59–72.

WILSON, G V, and FLETT, J S. 1921. The lead, zinc, copper and nickel ores of Scotland. Special Report on Mineral Resources. Memoir of the Geological Survey of Scotland, (Edinburgh: HMSO).

WINKLER, H G F. 1979. Petrogenesis of metamorphic rocks. (5th Edition. edition). (New York: Springer-Verlag.)

YARDLEY, B W D. 1978. Genesis of the Skegit Gneiss migmatites, Washington, and the distinction between possible mechanisms of migmatisation. Geological Society of America Bulletin, 89, 941–951.

Figures, plates and tables

Figures

(Figure 1) Summary solid geological map of the Ballater district.

(Figure 2) Topographical map of the district and environs.

(Figure 3) Bouguer gravity anomaly map for the Ballater district, contour values in mGal. Outline geology taken from the 1:250 000 Series map. Lineaments observed in images of the regional gravity (solid lines) and magnetic (dashed lines) data. Shaded region is where the total magnetic intensity anomaly exceeds 100 nT.

(Figure 4) Ground magnetic data in the Ballater district. Amplitude trace plot of the total magnetic intensity at 2 m above ground level after regional correction. Other ornament as (Figure 3).

(Figure 5) Geophysical model across the Ballater district. The profile shows the upper section of a whole-crust geophysical model of density and magnetisation which approximately models the observed regional gravity and magnetic data. Values in the section represent density in Mgm−3 and susceptibility in 10−3 SI.

(Figure 6) Geophysical model of the ground magnetic anomalies in the Coyles of Muick area. Polygons represent magnetic units in the Dalradian and mafic intrusions. Values in the section are the susceptibility in 10−3 SI

(Figure 7) Lithostratigraphical columns for the Dalradian rocks of the Ballater, Braemar and Pitlochry districts.

(Figure 8) Map of the south-east quadrant of the district showing the position of the north-west-trending lineament and (inset) its effect on the lithostratigraphy of the upper Argyll Group Dalradian.

(Figure 9) Map showing the distribution of Dalradian rocks in the district.

(Figure 10) Generalised vertical section of the Glen Girnock Calcareous Formation in the area of the type section, near Strathgirnock.

(Figure 11) Geochemical characteristics of Dalradian amphibolites in the district. a FeO-Na2O+K2O -MgO triangular diagram showing characteristic iron enrichment of tholeiitic rocks b Plot of TiO2 against Zr/ P2O5 showing close spatial relationship of amphibolites to alkaline/ tholeiitic field boundary c Plot of Na2O against CaO showing the control of bulk chemistry on amphibolite mineralogy. Ringed samples contain clinopyroxene porphyroblasts

(Figure 12) Lithostratigraphical correlation of Dalradian rocks across the Glen Doll Fault.

(Figure 13) Distribution of major intrusions in the district.

(Figure 14) Geochemical variation diagrams for syntectonic and late-tectonic granites in the Ballater district. Circles denote data from the Cairn Trench Granite, squares the Rough Craig Granite.

(Figure 15) Triangular plots of normative Qz–Or–Ab for syntectonic and late-tectonic granites in the Ballater district. Three phase cotectic line at 5 kb superimposed.

(Figure 16) Distribution of main structural elements in the district.

(Figure 17) Distribution of regional and thermal metamorphism in the district.

(Figure 18) Possible qualitative P—T time paths for regional metamorphism.

(Figure 19) Petrogenetic grid for Dalradian rocks within the thermal aureoles of post-tectonic granite and diorite bodies in the district.

(Figure 20) Geological sketch map of the Lochnagar Granite.

(Figure 21) Geochemical variation diagrams for post-tectonic intrusive rocks of the Abergeldie, Cul nan Gad, Glen Gairn, Juanjorge, Khantore and Lochnagar bodies based on new BGS analyses and data from Harrison (1987).

(Figure 22) Distribution of minor intrusions in the district.

Plates

Front cover Aerial photograph of Lochnagar showing north-facing corrie incised in Lochnagar Granite, and 1050 m erosion surface. (D5236) (Photographer T Bain)

(Plate 1) Contrasts in valley morphology between the northern and southern parts of the Ballater district. a The Dee valley at Crathie. View eastwards from Craig Nordie [NO 2352 9445] showing the wide nature of the valley and generally subdued topography. The area is underlain by Dalradian Appin Group metasediments which are extensively intruded by granitic and dioritic units of the Abergeldie Complex. Balmoral Castle (middle ground) and Crathie Church (background) are both built with Lochnagar L2 Granite. (D4252). b Upper Glen Clova. View north-westwards from near Bassies [NO 29 73] showing classic parabolic glaciated valley. The area is largely composed of Dalradian metasedimentary rocks of the Southern Highland Group, but the lower ground and the forested area in the middle distance underlain by the post-tectonic Glen Doll Diorite. The summit of Lochnagar is visible in the distance (P100665).

(Plate 2) Sculptured pillar of deeply weathered quartz-diorite on the eastern edge of the Abergeldie complex with quartzite scree behind. Location is in meltwater channel south of Buailteach [NO 2799 9309] (D4262).

(Plate 3) Disused limestone pit near the summit of Creag Phiobaidh [NO 3270 9472]. The limestone, seen at the northern end of the pit is about 3 m thick, dips at 45º to the south-east and is overlain by massively weathered calc-silicate rocks. The footwall is a 2 to 3 m thick felsite sheet. (P100662).

(Plate 4) Rusty weathering quartzite with interbedded laminated semipelite and quartz segregations, top of the Creag nam Ban Psammite Formation [NO 3138 9380]. The essentially flat surface shows an S-shaped F2 fold, the limbs of which are refolded by small upright F4 folds and the axial plane traces run from top left to bottom right. The hammer head is 165 mm long (D4260).

(Plate 5) Macroscopic and microscopic evidence of intraformational breccia in Glen Girnock Calcareous Formation on Creag Phiobaidh [NO 3277 9469]. a. Knobbly weathered surface reflecting coarse clastic nature of rock. The hammer shaft is 340 mm long (P100663). b. Photomicrograph of intraformational breccia showing fine-grained siliceous clast rimmed by garnet. (S94410), crossed nicols, magnification x 11 (MNS 10278).

(Plate 6) Deformed volcaniclastic breccia [NO 3133 9053] on Craig Megen. The photograph shows stretched clasts of hornblende-plagioclase rock in a matrix of hornblendic psammite. The hammer head is 165 mm long (D4257).

(Plate 7) F4 folds in gneissose semipelite, Queen's Hill Gneiss Formation [NO 3163 8975], Craig Megen. The hammer head is 165 mm long (D4261).

(Plate 8) Dounalt Limestone Formation in White Water, Glen Doll [NO 2335 7761], showing a sequence of generally thin banded impure limestone and calcareous psammite, duplicated in part by folding. Hammer shaft is 340 mm long. (D4266).

(Plate 9) Contact between sheared metagabbro and ultramafic rock (hornblendite) in Coyles of Muick syntectonic intrusion [NO 3241 3174]. The coin is 25 mm in diameter. (D4253).

(Plate 10) Strongly deformed older basic intrusion (top) cut by locally sheared rocks of the younger suite, Hill of Candacraig [NO 345 997]. The hammer head is 165 mm long. (D3772).

(Plate 11) Gneissose syntectonic granite on Rough Craig [NO 3543 7264] showing melanocratic (biotite rich) and leucocratic (quartz-plagioclase-microcline) segregations. The more mafic areas may represent digested xenoliths or 'restite' material. The hammer head is 165 mm long. (D4268).

(Plate 12) Second generation migmatite (oligoclasebiotite gneiss) with psammite inclusions containing early isoclinal fold, in Queen's Hill Gneiss Formation on Craig Megen [NO 3194 8921]. The hammer head is 165 mm long. (P100664).

(Plate 13) F3 recumbent folds in Green Beds, comprising biotite schist (left) and hornblende schist (right). The hammer head is 165 mm long (D4549).

(Plate 14) Photomicrographs of Dalradian metasedimentary rocks illustrating Barrovian metamorphic mineral assemblages and textures. a. Biotite porphyroblasts, magnification X 24 (MNS10289) b. Muscovite herringbone, magnification X 20 (MNS10290) c. Kyanite and staurolite, magnification X 24 (MNS10272) d. Kyanite in biotite-rich lithon, magnification X 24 (MNS10291) e. Fibrolitic sillimanite, magnification X 70 (MNS10292) f. Kyanite in fibrolite, magnification X 42.5 (MNS10294) g. Cross-cutting fibrolitic sillimanite, magnification X 70 (MNS10266) h. Fibrolitic sillimanite cut by prismatic sillimanite, magnification X 42.5 (MNS10263)

(Plate 15) Photomicrographs of Dalradian metasedimentary rocks, illustrating Buchan metamorphic mineral assemblages partly affected by thermal overprint. a Andalusite and fibrous sillimanite in hornfelsed aluminous semipelite, Birchwood Semipelite Formation, near Drummargettie [NO 2414 9558] (S77613) plane polarised light, magnification X 70 (MNS10259). b Possible regional andalusite in hornfelsed aluminous semipelite, Birchwood Semipelite Formation, near Drummargettie [NO 2414 9558] (S77613) plane polarised light, magnification X 11.5 (MNS10260). c Staurolite mantled by andalusite, cordierite and hercynite in schistose semipelite, Creag nam Ban Psammite Formation [NO 3134 9381] (S82068) crossed nicols, magnification X 70 (MNS10271)

(Plate 16) Intrusive relationships in the Abergeldie Diorite Complex. a Xenolithic granodiorite cutting porphyritic quartz-diorite ('mafic spot' rock) [NO 2694 9313].The hammer shaft is 340 mm long. (P100666). b Quartz-diorite enclosing xenoliths of feldspar-phyric syntectonic basic rock and coarse-grained diorite [NO 2701 9295], The hammer head is 165 mm long. (D4255).

(Plate 17) Photomicrographs of intrusive rocks of the Abergeldie Diorite Complex. a Medium-grained porphyritic quartz-diorite showing composite nature of 'mafic spots' (S92666) plane polarised light, magnification X 42.5 (MNS10275). b Coarse apatite and euhedral sphene in coarse-grained diorite (S92921) plane polarised light, magnification X 42.5 (MNS10264). c Relict clinopyroxene, largely overgrown by biotite and hornblende in granodiorite (S77631) plane polarised light, magnification X 70 (MNS10267). d Coarse zircon in granodiorite (S92661) plane polarised light, magnification X 70 (MNS10255). e Bladed biotite crystals in Creag nan Gall Granodiorite (S77637) plane polarised light, magnification X 42.5 (MNS10258) . f Coarse cloudy apatite in white granite (S92977) plane polarised light, magnification x 70 (MNS10269).

(Plate 18) Foliated, porphyritic and xenolithic granodiorite [NO 2890 9193] on Tom Bad a' Mhonaidh, cut by white Lochnagar L3 granite. The two lithologies are separated by a narrow vein of white aplite white also intrudes the granodiorite. The hammer head is 165 mm long (D4254).

(Plate 19) Quartz-feldspar porphyry dyke veined by Coilacriech Granite pegmatite. The lens cap is 55 mm in diameter. (P100667).

(Plate 20) Photomicrographs of Dalradian metasedimentary rocks showing thermal metamorphic mineral assemblages and textures. a Fresh cordierite with biotite inclusions (S82012) plane polarised light, magnification X 70 (MNS10256). b Cordierite and biotite replacing garnet, and cordierite after muscovite (S82067) crossed nicols, magnification X 11.5 (MNS10253). c Hercynite with euhedral outline in cordierite (S82068) plane polarised light, magnification X 70 (MNS10270). d Prismatic sillimanite (S92859) plane polarised light, magnification X 42.5 (MNS10261). e Hercynite and corundum in andalusite mantled by biotite and corundum (S82068) crossed nicols, magnification X 42.5 (MNS10265). f Prismatic sillimanite pseudomorphed by andalusite (S92675) crossed nicols, magnification X 42.5 (MNS10262).

Tables

(Table 1) Geological sequence of the Ballater district.

(Table 2) Magnetic susceptibility of selected rocks from the district in S I units.

(Table 3) Chemical analyses of cummingtonite-bearing rocks.

(Table 4) Representative modal analyses of post-tectonic diorites and granodiorites.

(Table 5) Representative modal analyses of post-tectonic granites.

(Table 6) Chemical analyses of post-tectonic diorites and granodiorites.

(Table 7) Chemical analyses of post-tectonic granites.

Tables

Table 2 Magnetic susceptibility of selected rocks from the district in S I units

Formation Minimum Maximum Mode Sample
Balnacraig Metabasite Member 2 x 10−4 8 x 10−4 7 x 10−4 6
Queen’s Hill Gneiss Formation
Psammites 8 x 10−5 6 x 10−2 5 x 10−4 47
Metabasites 1 x 10−4 6 x 10−2 5 x 10−3 56
Water of Tanar Limestone Formation 8 x 10−5 2 x 10−3 5 x 10−3 17
Tarfside Psammite Formation
Glen Tanar Quartzite Member 8 x 10−5 1 x 10−2 3 x 10−4 40
Cald Burn Gneiss Member 8 x 10−5 4 x 10−4 1 x 10−4 6
Longshank Gneiss Formation 8 x 10−4 2 3 10−2 9 x 10−3 9
Younger Basic rocks 8 x 10−4 6 x 10−5 5 x 10−5 9
Post-Caledonian Instrusive rocks
Diorite–tonalite 6 x 10−5 8 x 10−2 1 x 10−2 12
Granites 1 x 10−4 1 x 10−2 9 x 10−3 10
Metasedimentary hornfels 8 x 10−5 8 x 10−3 5 x 10−4 39
Tadesse, 1991
Glen Girnock Calcareous Formation 4 x 10−4 4 x 10−3
Metabasites 2 x 10−3 2 x 10−2
Queen’s Hill Gneiss Formation 5 x 10−4 2 x 10−3
Coyles of Muick ultramafic rocks
Sheared Gabbros 1 x 10−4 9 x 10−3
Gabbros 6 x 10−3 9 x 10−2
Serpentinites 2 x 10−2 9 x 10−1
Diorites 9 x 10−2 5 x 10−1
Granites 2 x 10−3 9 x 10−3

Table 3 Chemical analyses of cummingtonite-bearing rocks.

(S2284) (S2294)
SiO2 59.08 58.93
TiO2 1.28 0.84
Al2 O3 17.82 18.96
Fe2O3 7.57 5.96
MnO 0.32 0.27
MgO 7.16 5.88
CaO 1.02 2.43
Na2O 1.67 1.91
K2O 0.39 0.76
P2O5 1.4 0.95
Total 100.24 98.76
Ba 578 503
Ce 85 117
Co 14 21
Cr 82 150
Cu 9 14
Hf 5 6
La 43 58
Nb 22 35
Ni 34 63
Rb 79 87
Sc 16 26
Sr 465 374
V 86 130
Y 29 39
Zn 75 93
Zr 186 249

Table 4 Representative modal analyses of post-tectonic diorites and granodiorites

1 2 3 4 5 6 7 8 9
Quartz 16 10 12 13 6 17 15 16 12
Plagioclase 52 38 46 49 43 45 51 47 55
K-feldspar 2 7 5 0 0 5 1 10 17
Biotite 13 6 16 16 9 23 21 13 12
Hornblende 14 38* 19 20 39 9 11 12 1
Sphene 0 T 0.5 1 0 0 1 1 2
Opaque minerals 1 T 0.5 1 3 0 0 1 1
Apatite 0 T 0 0 0 0 0 0 0
Zircon 0 0 T 0 0 T T T T
  • * Includes possible relict clinopyroxene T Trace
  • 1 (S92864) Tonalite, Juanjorge
  • 2 (S82164) Quartz-monzodiorite, Cul nan Gad
  • 3 (S92666) Porphyritic medium-grained diorite, Abergeldie
  • 4 (S92665) Medium-grained diorite, Abergeldie
  • 5 (S92676) Coarse-grained diorite, Abergeldie
  • 6 (S92662) Quartz-diorite, Abergeldie
  • 7 (S92681) Tonalite, Abergeldie
  • 8 (S92685) Granodiorite, Abergeldie
  • 9 (S92661) Creag nan Gall type granodiorite, Abergeldie

Table 5 Representative modal analyses of post-tectonic granites

1 2 3 4 5 6 7 8 9
Quartz 22 23 29 25 45 34 33 34 32
Plagioclase 34 39 41 33 27 38 27 36 32
K-feldspar 32 28 26 40 26 14 39 27 33
Biotite 12 8 4 2 1 9 2 4 3
Hornblende 0 1 0 0 0 3 0 0 0
Muscovite 0 T 0 0 T 0 0 T T
Sphene 0 T T T T 1

TT

 T
Opaque minerals T T T T T T T T T
Apatite T T T T T T 00 0
Zircon T T 0 T 0 0 0 0 0
  • 1 (S92874) White granite, Abergeldie
  • 2 (S82510) L1 granite, Lochnagar
  • 3 (S82502) L2 granite, Lochnagar
  • 4 (S82505) L3 granite, Lochnagar
  • 5 (S82516) Microgranite, Lochnagar
  • 6 (S77661) Granite, Glen Gairn
  • 7 (S77680) Leucogranite, Glen Gairn
  • 8 (S82014) Granite, Coilacriech
  • 9 (S82296) Granite, Ballater T Trace

Table 6 Chemical analyses of post-tectonic diorites and granodiorites

% 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19
SiO2 51.20 52.86 53.78 54.55 55.39 56.17 56.53 56.84 52.23 57.31 57.36 57.63 57.67 62.17 62.62 63.03 63.14 64.25 67.28
TiO2 1.55 1.60 1.35 1.41 1.30 1.61 1.22 1.28 1.19 1.10 1.16 1.19 1.13 0.86 0.90 0.83 1.13 0.82 0.58
Al2O3 16.43 16.30 17.12 17.51 17.50 16.69 17.71 17.83 15.51 15.16 15.44 17.37 15.92 16.46 17.17 17.38 16.31 17.71 16.32
Fe2O3 8.94 8.67 7.20 7.37 7.34 7.37 6.75 6.64 6.97 6.67 6.98 6.53 6.40 4.33 4.61 4.22 4.97 3.83 3.07
MnO 0.19 0.18 0.16 0.14 0.15 0.16 0.15 0.15 0.17 0.15 0.16 0.15 0.14 0.12 0.12 0.12 0.12 0.09 0.10
MgO 6.48 6.09 5.02 4.14 3.95 3.95 3.67 3.16 5.40 6.45 5.26 3.21 4.94 1.71 1.67 1.52 1.77 1.24 0.95
CaO 8.81 7.80 8.51 6.76 6.22 5.87 5.50 5.48 5.61 5.30 5.51 5.16 5.22 3.43 3.93 3.42 3.78 3.05 1.66
Na2O 2.82 3.02 3.49 4.11 3.85 3.80 3.87 4.04 3.26 3.03 2.70 4.00 3.91 4.32 4.31 4.51 4.17 4.67 3.54
K2O 1.49 1.59 1.35 1.77 1.96 2.42 2.12 2.49 2.28 2.43 2.73 2.66 2.52 2.97 3.00 3.83 3.09 3.37 4.97
P2O5 0.37 0.20 0.39 0.46 0.36 0.32 0.39 0.38 0.31 0.33 0.19 0.31 0.25 0.30 0.24 0.20 0.33 0.24 0.20
LOI 1.21 1.04 1.02 1.15 1.20 0.80 1.71 1.02 1.57 1.33 1.46 1.01 1.06 2.72 0.70 0.50 0.55 0.47 1.04
Total 99.52 99.36 99.39 99.38 99.23 99.16 99.63 99.31 99.52 99.26 98.97 99.23 99.19 99.39 99.27 99.56 99.36 99.74 99.71
Ba 406 332 379 590 553 438 620 868 448 453 926 766 450 946 631 1422 714 959 1417
Ce 46 45 52 98 71 58 73 70 67 46 42 68 62 80 71 108 98 81 111
Co 26 31 22 19 19 23 17 15 24 30 26 17 23 8 10 8 11 4 5
Cr 147 96 62 54 73 38 63 37 123 274 144 57 161 22 13 15 15 5 12
Cu 22 26 26 21 21 19 23 16 18 25 15 17 26 7 8 9 9 7 5
La 19 17 31 45 31 25 29 28 26 23 18 30 28 38 32 38 53 42 52
Nb 10 11 12 15 14 15 12 13 14 12 11 15 12 16 14 20 19 14 11
Ni 40 45 19 29 28 20 23 18 55 116 70 23 70 8 5 5 6 2 3
Pb 7 6 7 6 10 11 5 11 6 9 10 8 8 16 18 17 13 17 15
Rb 42 41 27 51 45 73 55 76 70 61 92 73 67 70 100 82 80 106 107
Sr 635 649 784 734 747 601 747 948 568 605 612 697 568 631 527 660 488 552 496
Th 4 3 5 6 4 7 6 7 4 6 5 6 7 9 9 10 15 12 12
U 3 3 5 5 1 4 4 3 2 3 2 5 4 3 5 3 6 5 5
V 161 166 136 134 119 113 110 102 98 111 120 101 105 64 71 57 72 36 37
Y 18 17 19 21 26 21 24 21 16 16 15 24 19 21 20 32 344 19 16
Zn 79 74 66 68 85 67 78 80 81 60 70 81 68 62 72 61 64 65 35
Zr 121 141 208 256 232 258 24 204 53 168 149 329 213 293 308 335 344 480 337
  • Key to analyses
  • 1 (S92873) Coarse-grained diorite, Abergeldie
  • 2 (S92988) Medium-grained diorite, Abergeldie
  • 3 (S82167) Diorite, Cul nan Gad
  • 4 (S92976) Tonalite, Abergeldie
  • 5 (S92864) Tonalite, Juanjorge
  • 6 (S92987) Tonalite, Abergeldie
  • 7 (S92865) Tonalite, Juanjorge
  • 8 (S92967) Tonalite, Glen Gairn
  • 10 (S92974) Quartz-diorite, Abergeldie
  • 11 (S92966) Tonalite, Abergeldie
  • 12 (S92872) Quartz-diorite, Abergeldie
  • 13 (S92959) Granodiorite, Abergeldie
  • 14 (S77642) Hybrid granodiorite, Abergeldie
  • 15 (S92990) Granodiorite, Creag nan Gall
  • 16 (S82844) Granodiorite, Abergeldie
  • 17 (S77637) Creag nan Gall type granodiorite, Abergeldie
  • LOI Loss on Ignition

Table 7 Chemical analyses of post-tectonic granites

% 1 2 3 4 5 6 7 8 9 10
SiO2 70.50 70.72 71.07 71.96 73.24 73.77 75.15 75.71 75.91 76.34
TiO2 0.42 0.43 0.36 0.34 0.29 0.19 0.18 0.12 0.11 0.09
Al2O5 15.00 14.92 14.97 14.43 14.54 13.33 12.96 12.67 12.32 12.56
Fe2O3 2.46 2.20 1.93 1.85 1.77 1.11 1.16 0.86 0.64 0.65
MnO 0.10 0.09 0.67 0.08 0.08 0.08 0.08 0.07 0.07 0.08
MgO 0.95 0.63 0.51 0.56 0.61 0.13 0.29 0.07 0.06 0.05
CaO 1.93 1.78 1.21 1.36 1.74 0.63 0.86 0.43 0.36 0.40
Na2O 3.83 3.93 3.33 3.72 3.78 3.53 3.58 3.51 3.71 3.74
K2O 4.13 4.17 5.27 4.33 4.14 5.31 4.32 5.07 4.32 4.66
P2O5 0.00 0.21 0.10 0.08 0.05 0.00 0.04 0.00 0.00 0.00
LOI 0.69 0.39 0.51 0.60 0.41 0.22 0.37 0.25 0.25 0.30
Total 100.01 99.47 99.33 99.32 100.65 98.30 99.00 98.76 97.94 98.87
Ba 560 552 802 503 438 641 170 21 62 10
Ce 73 64 94 66 55 96 35 66 36 36
Co 5 4 1 2 3 0 1 0 0 0
Cr 26 15 10 18 21 16 18 10 11 9
Cu 4 3 2 2 3 2 3 1 2 1
La 31 31 48 32 23 37 17 24 7 9
Nb 14 13 10 15 12 13 12 15 19 15
Ni 5 2 2 3 4 0 1 0 0 0
Pb 21 25 26 27 29 29 27 25 24 19
Rb 132 133 126 155 171 138 147 159 149 124
Sr 305 265 272 231 239 96 138 35 27 9
Th 18 15 12 19 17 19 23 18 17 13
U 5 4 6 7 8 5 6 8 6 3
V 30 24 19 20 20 4 11 2 3 2
Y 15 9 11 14 10 11 7 13 6 9
Zn 42 39 30 35 36 27 22 25 43 23
Zr 188 202 211 165 136 176 98 116 74 74
  • 1 N2746 L1 granite, Lochnagar
  • 2 (S92989) Monzogranite, Abergeldie
  • 3 (S92977) White granite, Abergeldie
  • 4 (S82523) Microgranite, Lochnagar
  • 5 (S82517) L2 granite, Lochnagar
  • 6 (S92991) L3 granite, Lochnagar
  • 7 (S92968) Leucogranite, Glen Gairn
  • 8 (S92985) L3 granite, Lochnagar
  • 9 (S92961) Granite, Khantore
  • 10 (S82526) L3 granite, Lochnagar