PAPERS AND PROCEEDINGS OF THg ROYAL SOCIETY . By JAMES A. PETERSON Department oj Geography, Monash Universit:!J (With four plates and four text figures.) ABSTRACT Cap summit. The peaks contained within the mountain area form a continuous range from the Mapping of glacial landforms of the French Clytemnaestra north to the Frenchmans Cap and mans Cap National Park shows that an independent the Lion's Head, then east by the razorback ridges system of cirque and valley glaciers accumulated at the head of the Livingston Valley, to Nicoles upon and caused considerable modification of this Needle and the Philps Peak massif. These moun area during the last Pleistocene glaciation. Clear tains are formed of the resistant siliceous Pre est evidence of the sequence of advance and retreat cambrian metamorphics of the west and south is seen in the Vera Valley where, during the maxi west of Tasmania, standing locally as ridges and mum phase, the ice reached the Vera Creek gorge peaks above the steep-sided valleys, which here and overspilled a series of diffluence passes further are cut into softer schists and phyllites. To the upstream. The history of subsequent retreat can east the quartzites drop away through dense be traced with reference to the post-maximum temperate jungle to the Loddon Plains. These moraines. button-grass plains are base-levelled to rock-bars Some detailed work on the morphology of the at the heads of gorges that eventually lead to the lake basins makes possible a classification of lakes Franklin River, the master stream of the region. which shows clearly the nature of the glaciation Marked by its incised character and structurally and the influence of lithology. Careful considera controlled sections and rapids, this river drops tion of the aspect, morphology, and elevation of from 950 feet near the confluence with the Loddon the cirques in which many of these lakes occur to 550 feet at the confluence with the Livingston provides interesting comment upon the factors 25 miles downstream. The Cap area drains to governing accumulation and ablation in the mari the Franklin River in a more or less radial pattern, time mountain environment of glacial and late in places closely controlled by strike and fault glacial times. directions. The area has a wet, cool, temperate Radiocarbon assay of leaf fragments in rock-flour climate and is largely covered with rain forest, below peat in a bedrock depression on the lip of which makes access to much nf it a matter of one of the high discrete cirques gives a minimum difficulty. age of 8720 ±220 years B.P. for the final deglacia No previous geomorphological work as such has tion of the Frenchmans Cap area in particular and been completed in the area although the physio the western Tasmanian mountains in generaL graphic problems of nearby areas to the north The present study is discussed with reference to have been discussed by Spry (see below). Earlier some of the chronological problems of the last publications include papers by Ward (1908a, 1908b, glruciation. The possibility of a sequence of glacials 1909) and papers by Finucane and Blake (1933), and interglacials is recognized. As the Pleistocene Blake ( 1936, 1937), and Ahmad, Bartlett and Green climate changes of the southern hemisphere form 0959). The only other references of note are the the broad context of the present study attention is accounts of early explorers, e.g., Franklin, Philp drawn to analogies in the New Zealand sequence. and Rumney, and Calder. Geology INTRODUCTION The major contributions to the geology of the The Frenchmans Cap area surrounds the prom area are the papers by Spry 0957), Spry and inent peak of that name in the west of Tasmania Zimmerman 0959), Spry 0962, 1963) and Spry some 25 miles east of Queenstown. It is bordered and Gee 0964). The first two accounts cover the to the west and north by the great bend of the geology of the nearby Mt. Mary and Mt. Mullens :P'ranklin River, while to the south and east lie areas respectively, while the last three concern the Lachlan and Loddon Plains. The present study, the Frenchmans Cap area (see map in Spry 1962, based on some 60 days field work. and on air photo p. 112). interpretation, covers an area of about 70 square Spry describes the Precambrian metamorphics miles. Complete air photo coverage exists as well of the area as " strongly deformed metasediments as maps at a scale of one mile to the inch (French with a low to moderate grade of regional meta mans Cap National Park map and Franklin sheet), morphism" (1962, P. 112). The gap area is mapped with contour intervals of 100 and 50 feet respec as Mary Group east of Franklin Group although tively. Spry has recognised the presence of Scotchfire The land varies in altitude from about 550 feet Group rocks 0963, p. 113). Further work in the in the gorge of the Franklin to 4739 feet at the area (Duncan. pers. comm.) shows the following:-· 117 118 GLACIATION OF THE FRENCHMANS CAP NATIONAL PARK Franklin Group Mica, garnet and albite schists with These rocks comprise the lower quartzites and amphibolites. dissected area west of the Cap. Mary Group Massive and schistose quartzites. Comprise the main ridges and The massive quartzites are white, peaks of the Cap area. thickly bedded and concordant. Scotchfire group Phyllites quartz-schists and dolo Generally underlying the ridges mitic schists. The phyllites are of Mary Group rocks and softer than the Mary Group and forming the saddles and have marked cleavage. gullies. This succession is a structural one and carries wall sapping in the comb ridge extending from the no other implications. As in the adjacent Pre Nicoles Needle group eastwards to Philps Peak. cambrian areas stratigraphic position and regional The erratic blocks on the crest of the pass are metamorphism are debatable (Spry 1962, p. 113). much more likely to be protalus than moraine left Detailed mapping is often difficult due to struc by transf!uence across the pass. The exact limits tural complexity, strong faulting, and differential of inundation of the ridges is hard to assess from silicification which causes the rapid and irregular air photos because the overridden appearance of changes in lithology. There have been at least the high schist outcrops is due mainly to the two periods of folding. Superimposed upon the geological structure. The hanging Marilyn valley effects of the Frenchman Orogeny is that of the has been selectively eroded out of dolomite. Below later (Devonian) Tabberaberan Orogeny: it is the 1000 foot backwall topped by Mary Group the evidence related to the latter folding that is quartzites and quartz-schist, Lake Marilyn, which mapped by Spry 0962, fig. 4, p. 113). is ,60 feet deep, is contained behind a rock lip O'f dolomite. Down valley a second and much smaller The influence of lithology on the topography has lake, now rapidly being filled by colonizing vegeta already been outlined. The influence of structure tion, hangs above a 400 feet drop to Lake Vera. on topography is evident in the drainage pattern. Here, at the meeting point of the Marilyn and The area is in that part of Tasmania where " the Barron glaciers, the half-mile long lake is 180 feet folded Precambrian basement is uncovered and deep and very much "down at heel" (see Fig. 3). the rivers have evolved a trellis pattern with parallel ranges marking the strike of the harder rocks " The nature of the preglacial valley into which (Davies in Gill, 1962, p. 244). This is true of the the glacier expanded has caused an interesting Frenchmans Gap area as a whole and of the Frank series of features. Like the Tahune valley to the lin Group rocks to the west in particular. The west, the upper section is joint controlled. Both parts immediately surrounding the Cap itself show the Tahune and Vera valleys turn sharply to follow the influence of fault or joint control in the align the strike direction in their middle sections. Unlike ment of the valleys. the Tahune glacier, the Vera glacier was severely restricted where its valley changed direction, w GLACIAL GEOIU:ORPHOLOGY that the low strike ridges on the outside of the It seems reasonable to suppose that the gladers bend were overridden as the ice spilled out through of the Frenchmans Cap area were superimposed on five diffluence passes overlooking the Loddon Plains. a fluvial landscape of dominantly joint- and fault Although there is no evidence of how far these controlled valleys interrupted by small areas of outlet glaciers advanced towards the plains, they karstic depression some of which are now occupied were respons1ble for the agglomeration of tillflow by glacial lakes. The resulting glacial landforms (Hartshorn 1958-footnote p. 477, Sharp 1960, p. are typical of mountain glaciation in the high 333) and poorly stratified outwash, the coarse lands of Tasmania. Catenary troughs, cirques, nature of which is explained by the comparatively cirque-lakes, comb ridge aretes and moraine ridges high gradient. Below these deposits, valley trains are common features above 2000 feet, while out were laid down on the plains as may be seen where wash valley trains oocur at lower elevations (Fig. Philps Creek has incised five feet into outwash 1). Bas~cally a complex and high cirques and immediately upon reaching the flatter gradients radial valleys, the area contains eight cirque and of the Loddon Plains. The aerial photographs valley glacier systems heading at altitudes between suggest that this valley train was braided. 3000 feet and 3400 feet, with lengths varying between three miles and less than half a mile, and The weathered state of the schist boulders in the areas from eight square miles to less than one till on the flanks of the hills separating the quarter of a square mile. The following descrip diffluence ice tongues above the upper Vera Plains tion should be read with reference to Figs. 1 and indicates that the deposits are among the oldest 2 and one of the published topographic maps. surficial sediments in the region. The lower Vera valley itself provided a further The Vera glacier outlet. It was considerably modified, the pre Perhaps the most interesting valley is that once glacial gorge-like walls being overridden and a occupied by the Vera glacier and its tributaries. giant roche moutonnee formed at the valley bend, Fed from two cirques in the Philps Peak massif, the walls of which exhibit vertical stoss and lee the Marilyn and the Barron, this glacier left evi features (Lewis 1949). Ice from two north bank dence of expansion to a valley glacier maximum tributary cirques heading at the comb ridge west stage and subsequent retreat to the cirques after of Nicoles Needle may have joined the lower Vera a post-maximum still-stand. glacier at this stage. The larger of the two cirques heads at a col The maximum extent of the lower Vera glacier known as the Barron Pass, a gap opened by head- coincided with the Vera Creek gorge. Drift fills FRENCHMANS CAP NATIONAl PARK FIG. 1.-Location I\-1ap. Frenchmar:s Cap National Park {No.menc1ature fron1 the National Park Map and from the climbing notes of J. l\1:iddleton-Elliott). 1;z 0 1 L_.._,_.J__,.;.._..L...,.______j MIL. ES 1 y LOCAL LIMITS OF .or.,...,...... CIRQUE c· o" 1- y ICE ACTION ~( ~~\ . ~ ~ ~ .;. TRANSFLUENCE COL )( ROCK BASIN o-- DIFFLUENCE COL ...... MELTWATER CHANNEL ~ -/1\111\ \_~/ MORAl NE Rl DGE -=- MAMILLATED SURFACE GLACIFLUVIAL DEPOSIT cJ- ABRADED VALLEY STEP r-1 / GROUND MORAINE ~:~:-:~:> l I F'm. 2.-Gla.cial Landforms of the F1~enchmans Cap massif. JAMES A. PETERSON 121 the lower half of the valley above the gorge and FIG. 3.----Isobathic maps Lake Vera and Lake Whitham. .has been terraced by postglacial river erosion, the present Vera Creek being incised into the till. In places the till extends to some 50 feet above the ~-----·------·----· river level and must be at least 30 feet thick. It contains a large percentage of boulders, is coarse, and permeable, with a matrix of pro I VERA nounced sandiness. The glacier snout may have extended a short distance into the gorge, but, as would be expected, there is no evidence for this. The narrow (20 feet to 30 feet at the base) 200-300 .feet high gorge marks a distinct change of character Isobath Interval 10 feet in the valley. Above the gorge are glacial deposits; below, the valley is deeply incised and V-shaped. The same is true of the Tahune valley to the west. That glaciers from the Frenchmans Cap area were responsible for Pleistocene damming of the Frank lin valley as suggested by Spry and Zimmermann (1959), to explain the presence of varved clays in the Artists Hill-Mt. Mullens area upstream is not immediately apparent. At the glacial maximum ice in the upper Frank lin and adjacent valleys may well have been coter minous with the ice of the St. Clair district and ipso tacto with that of the central plateau area. Ice tongues from such a huge accumulation area would be pushed long distances. Ice in the Cap area on the other hand accumulated in a strictly limited area, from whi,ch the ice soon moved to the lower levels of the ablation zone. It is suggested o YARDS 220 then that although evidence is lacking it is most [ ____L______j likely that the same ice which deposited till on Artists Hill and left large erratics at about the same height near the foot track below, and north of, the Mt. Mullens saddle, might well have placed a mor- aine dam downstream. The maximum extent reached by outer lobes o.f the main central ice cap of Tasmania is still a matter of debate. That glacial advance into the Franklin-Gollingwood river junction area was a maximum stage feature is WHITHAM illustrated by the weathered state of the Artists Hill till (Spry and Zimmermann 1959, p. 3). It is important to realise that the location of this till is a better indication of its age than is the weathering state of dolerite boulders contained therein. The amount of weathering of dolerite till stones in a given time will vary from place to place due to differences in composition of the dolerite, composition of the till matrix and relation to ground water fluctuation. Moreover there is the possi bility that drift containing weathered dolerite was transported en-masse, probably in the .frozen state. This process has been suggested as a possible explanation for the presence of fragile lumps of Permo-Triassic sandstones in the bedded moraines at Lake St. Clair (Derbyshire 1965, p. 381). Perhaps these factors explain the anomalies of the weather Isobath Interval ing state of different tills of probable similar age, 20 feet e.g., the Arm River till and the Artists Hill till (Spry 1958, Spry and Zimmermann loc. cit.). The next marked stage in the history of the Vera glacier is indicated by the large recessional moraine at Lake Vera. This rises steeply 350 feet above the lake outlet and 200 feet above the Upper Vera Plains. A cover of ground moraine and l~ scattered erratic boulders was left between the lake and the diffluence passes from which the glacier o YARDS 220 retreated. The Lake Vera moraine is aligned L..----~---...J north-south while the glacier responsible for its 122 GLACIATION OF THE FRENCHMANS CAP NATIONAL PARK formation came from the southwest. This align This general lack of cirque lip moraines is ment was probably governed by the rate of retreat interesting. On the one hand it suggests that the of the various parts of the ice front from the rock basins contained active ice mainly during the diffluence passes (for some of the ice tongues had glacial maximum (see Dart 1957). The lack of spread out more acutely to the main direction of evidence for re-occupation of overridden cirques flow than others), and also by the position of strike after the retreat of the Central Plateau ice cap ridges which outcrop at several places on the (Jennings and Ahmad 1957) may be taken as northern end of the moraine. The retreat of the support. for this. Certainly rotational slip of cirque glacier at this stage can be traced in detail. glacier ice, postulated as the cirque forming agent The differential withdrawal from the maximum elsewhere (Lewis 1960), is, following the calcula extent at the diffluence passes was most rapid from tions of Lewis 0949), much more lil NANCY JETTY GWENDOLEN RUTH !------..------+- MARYLIN TAHUNE 0 YARDS 220 Isobath Interval /0 feet '------~-__j FIG. 4.--Isobathie Maps of the smaller cirque lakes. 124 GLAClATION OF THE FRENCHMANS CAP NATIONAL PARK this glacier it is justifiable to postulate penetration looked by the west and north cols of the Cap. A of the lower valley with only limited maximum lake basin, 90 feet deep, was eroded in massive stage transfiuence of Deception Ridge. quartzite and quartz-schist under the deepest part A more detailed history of the valley would of the glacier, which, judging by the height of the require lake sounding. The main lakes were too Nancy col on the northern side of the trough, was large for accurate single-handed work with sound over 400 feet thick at its maximum stage. At this ing line and the depth sounder used was incapable tirne both the Sophie and Gwendolen glaciers were of receiving signals from the muddy lake bottoms. cascading down V -shaped valleys below their The following comments are therefore tentative. troughs, and melting rapidly among the outwash Lake Millicent lies immediately upstream of the deposits, till and protalus. deposited during the valley constriction and is probably partly moraine glacial advance to this maximum. These deposits dammed. Lake Magdalen also is moraine dammed. are braided below the Gwendolen trough and several However its situation in the widest part of the larger abandoned channels can be seen below the valley below a mammilated valley step which forms Sophie trough. Glacial deposits have buried low the low dified northern shoreline of the lake ridges of the softer Franklin Group rocks between suggests it is a partial rock basin. Lake Gertrude, the Gwendolen and Ness Creeks. situated below a small cirque below the junction Subsequently the drift deposits at the end of the of softer Scotchfi.re and the Mary Group rocks Gwendolen trough were laid down. This material has swampy shores on deep clay and silt detritus. is predominantly composed of gravel and boulders The lake has been subject to infilling from both with sandy fines. Under the backwall (which is the Cecily and Gertrude cirques. The small cirques not as uniformly vertical as it might have been on the west wall of the trough probably represent had the geological structure been favourable) lee the late cirque stage seen in the other valleys. ends of well formed roches moutonnees project into the lake. The Sophie and Upper Cwm glaciers The trough of the Sophie glacier is located on The Tahune glacier the west side of the Clytemnaestra Ridge. This On the north side of the Cap the Tahune glacier glacier was substantially augmented by ice from headed beneath a steep backwall rising from 2200 the Upper Cwm situated under the Cap summit to 4700 feet. This was broken at the 2700 foot on the west-southwest side. This ice may have level by the small, high cirque containing the rock been coterminous with the ice of the Gwendolen basin of Tahune. The main valley, as men glacier (via the west coD and the Cecily glacier tioned above, is probably joint or fault controlled (via the south col), at the glacial maximum. It in its upper sedion, and strike controlled in descended to join the Sophie glacier over two steps, its lower section above the T'ahune gorge. At between which lies the small rock-basin tarn, Lake the maximum extent slight diffluence occurred from Charlotte. Some of the finest mammilated rock the outer part of the valley bend, while the glacial surfaces to be found in Tasmania occur in the snout reached at least one and a quarter miles to Upper Gwm. Smooth stoss and lee profiles in the the Tahune gorge. As with the Vera valley there roches moutonnees reflect north-south movement is a distinct change of character of the river valley although inherent weaknesses in the quartz-schist at the gorge. There is also a eontrast between the bedrock have been exploited regardless of ice move upper and lower sections of the glaciated part of ment direction as seen by the small, comparatively the valley. The broad catenary shape of the upper deep, rock basins (see Plate 2). section shows that it was more susceptible to modi Lake Sophie is a shallow lake (30 feet deep at fication than the strike valley section, despite the most), dammed by a relatively extensive deposit additional contributions of ice from the Nancy of drift in the form of a series of recessional and Jetty Lake areas to this part of the valley. moraines. A section ten feet deep, exposed by the These cirques were bounded to the east by the lake outlet, shows a bouldery till of quartzite, upper Tahune ice while they built small lateral quartz-schist, and chlorite schist boulders and moraines on their western sides. Air photo inter cobbles in a sandy matrix. The lateral moraine pretation suggests strongly that the position of standing some 200 feet above the southern edge of these cirques was governed by what is taken to be the lake and running parallel to it is particularly a well marked structural weakness there. The drop well formed. Westward (downstream) it curves of 650 feet between the two eirques would have more and more sharply towards the lake, but is not accommodated a cascading glacier. Lake depths, coterminous with the recessional moraines. rock lips, and the mammilated bedrock of the plat form surrounding Jetty Lake indicate similar abra Further south between the Clytemnaestra and sive action to that already described in the Upper Mt. Electra a large west-facing cirque para1lels Cwm. In neither of these places were any striations the Sophie valley. It is a comparatively poorly found. Other less spectacularly abraded areas like developed form which, nevertheless, probably over wise lacked such marks of ice movement direction. topped the ridge to the southeast to become coter The conclusion is reached that these metamorphic minous with the southernmost cirque overlooking rocks are characterised by so many structural the Livingston Valley. Cons'dering the position of foliations that it is impossible to separate lineations these cirques it is not unlikely that snow patches due to glaciation from those resulting from the developed on the Deception Range in the lee of surface expression of minor structural features. Mt. Lyne (3050') several miles to the S.S.W. The Tahune cirque hangs nearly 1000 feet above The Gwendolen glacier the Tahune trough headwall. It is a small steep sided and overdeepened cirque excavated in Scotch Parallel to the Sophie glacier, and to the north, fire dolomite under the north face of the Cap. the Gwendolen glacier occupied the trough over- Lake Tahune, although little more than 100 yards JAMES A. PETERSON 125 across is over 70 feet deep and is contained behind very fine particles are carried away by meltwater a dolomite rock lip through which it drains by in areas of mountain glaciation. Tests for lime way cf solution channels. The rock-lip has a thin content of showed that any calcareous cap of bouldery till and protalus. materials had leached out, a condition The Whitharn glacier in this area. The results in fact were of of tills derived from con- Lake Whitham lies at 3060 feet in an over rock.,s vvhich show a high deepened cirque southeast of Philps Peak. Out proportion of boulders and pebbles a matrix of cropping dolomite at several places around the pronounced sandiness. lake shore indicates selective erosion of Scotchfire dolomite as in the Lake Marilyn cirque. The maxi The proportions of quartzite and quartz- mum extent of the Whitham glader is unknown makes dHl'erent:lation based 1n for the ice cascaded steeply after passing over the difficult as remain almost glacially moulded quartz-schist rock-lip. This lip as boulders in surftdal is well smoothed to a height of at least 70 feet deposits. a the predominantly quartzite till of the lower Vera valley shows very little above the lake level (Plate 3). of weathering, while tbe Vera diffluence The East Philp glaciers which may reasonably be assumed to be similar only other glaciers of the area were the age, contain weathered schist cobbles and boulders. shadow " cirques hlgh on the In contrast to quartzite, dolomite boulders may eastern of Philps Peak massif overlooking become completely decomposed as till the Loddon Plains. One of these cascaded on to found in pits excavated in the Vera the southernmost diffluence tongue of the Vera moraines. glacier. The other was situated on the east side In considering the significance of weathered till of Mt. Agamemnon. Neither were overdeepened particles the question arises as to how much drift enough to contain lakes and much of the material is derived from the preglacial weathered mantle deposited in them is not till but protalus, includ~ and how much is derived from the mechanical ing protalus ramparts. As with two other cirque action of ice on bedrock. While the ability of ice like features nearby they would have held only to abrade and erode unweathered bedrock is clearly snow-patches for most of the last glacial period. evident in the powerful glacial moulding of stoss and lee features, there is little evidence on which GLACIAL DEPOSITS to base an assessment of the amount of material Thirteen till samples were analysed by sieving, contributed to the drift by pre-glacial weathering. decantation and hydrometer analysis in the The weathered dolomite boulders mentioned above C.S.I.H.O. Soil Laboratory in Hobart. The samples almost certainly decayed after emplacement in brought out for analysis vary in size from 40 lbs. the till as glacial transport in mountain valleys to 2-3 lbs., the boulders having been measured in would be unlikely to allow the retention of cohesion the field. The small clay content of most samples in weathered boulders. The preglacial weathering is explained by the fact that most of the bedrock of the till stones in question is rendered further from which the tills were derived comprises resist unlikely by their presence in post-maximum stage ant quartzites and quartz-schists, and that most tills. GLACIAL LAKES A 'Classification of lakes (Table ll of the Frenchmans Cap area is useful in illustrating the glaciation of the area. It should be used in conjunction with a topographic map and all figures. TABLE I Type Name Remarks A. Cirque lakes due mainly to Whitham I Definite roxen lakes glacial erosion and over Marilyn f all selec.tively eroded i.n Scotchfire Group deepening. Tahune I rocks wi.th a high or 100% dolomite content. Other rock basins Nancy, Jetty* eroded in quartz-schists and quartzites Gwendolen a valley head cirque with a low partial moraine dam. Cecily partially moraine dammf:d-deeply eroded in softer Scotchfire Group rocks Sombre Tarn small cirque lake on valley side now almost filled by sedimentation and protalus. B. Valley lakes of compound Vera glacial overdeepening in Scotchfire rocks at origin dl!e to erosion and the junction of two glaciers. deposition. Sophie valley " cirque " lake--shallowed by debris from higher tributary cirque with greater accumulation area. Gertrude I valley lakes possibly marking retreat stage of l Magdalen I Livingston glacier. Lake Millicent may be Millicent I a miniature glint lake (see below). Other Marsh Lake I Small rock basins on steps below cirques. Lake Charlotte l *Jetty Lake is eroded in Srotchfire quartz-schist and mi;;ht Ce classed as a roxen lake. 126 GLACIATION OF THE FRENCHMANS CAP NATIONAL PARK Lake Millicent is thought to fit Hutchinson's type nices form on the Clytemnaestra Ridge during the 29 0957, p. 160), in that it is probably a rock-basin present winters and have been known to last regu resulting from "glacial corrasion in specific sites larly into October. They might reasonably be where the ice fiow is impeded by the pre-existing assumed to have been semi-permanent in late topography", although the examples quoted (op. glacial and early post-glacial times. cit. pp. 68 and 160), are on a much larger scale than the one in question here. Attention is drawn MULTIPLE GLACIATION to the different use of the term " glint lake " by The cirques can be divided into two groups on Charlesworth Cl957, pp. 250, 265, 285). The term the basis of altitude. Half those mapped in Fig. 2 "roxen lakes" (Charlesworth 1957, p. 265), has fall between 2100 feet and 2300 feet and the other been employed here for lakes selectively eroded in half between 2700 feet and 3400 feet. Early work the dolomite lenses of the Scotchfire Group rocl{S. in Tasmania used cirque levels to differentiate It is possible that some other rock basins were between the separate glaciations (Lewis 1922, 1924, initiated in depressions in dolomite which no longer 1933). It is now recognised (for example, Jennings outcrops. and Banks 1958) that there is no reason why cirques r.rhe isobathic maps (Figs. 3 and 4) were plotted at different altitudes cannot, contain ice at the from soundings made with the aid of a rubber boat same time, and, if they do not, then it is the high and sounding line. Plotting was done from bearings cirques that contain ice as a result of retreat of taken from the boat and from shore. They show ice from all lower levels. Gladal features in these that in some cases considerable overdeepening has late stage cirques will therefore be younger than occurred, especially in the high rock basins, most, features down valley. of which have their deepest points nearer the Cirque levels have also been used to determine backwall than the lip, which is of course typical orographic snowlines within different stages of of cirque lakes in general. the one glaciation. However, given the altitude necessary for the existence of solid precipitation the ASPECT AND MORPHOLOGY exact siting and level of any cirque is pre-deter Although the gla.ciation of the area was imposed mined by factors of accumulation (aspect, preci on a small, somewhat isolated massif, with a gener pitation amount and slope angle) and lithology. ally radial drainage pattern, the effects of advant Obviously to assume different ages for cirques on ages possessed by south and east facing valleys the basis of their altitudes is only an introductory and cirques for the accumulation of glacial ice can exercise to more detailed study. The " roxen lake " be seen. These advantages are two in number. cirques and the small high East Philp cirques of First, in areas of prevailing westerly winds, south the Frenchmans Cap area illustrate the effect of and east facing valleys are likely to accumulate local conditions on cirque glaciation. more snow due to both orographic precipitation Nor do the glacial deposits, which are in general and snow drifting. However, this advantage is a more promising indicator, provide any evidence also possessed by north facing valleys in the lee of multiple glaciation. Higher Wls are in some of north-south ridges. Secondly, south and east cases obviously less weathered than tills in lower facing slopes are more protected from insolation areas. This indicates nothing other than the usual (see also Charlesworth 1957, p. 298). It is signi single phase sequence of cirque glaciers expand ficant that the largest glaciers accumulated in the ing to a maximum as valley glaciers, followed by lee of north-south ridges while the largest cirque retreat into the original cirques. No signifcant faced south-east on the lee of the Cap itself. differentiation of the ftuvioglacial sediments is Closely related to this effect is the asymmetry evident. Although the evidence is consistent with of form shown mainly by the north-south ridges, in with only one glaciation it is well to remember that particular the Clytemnaestra Ridge including the proof of multiple .glaciation in Tasmania may yet Cap (Plate 4). As previously observed by A. N. Lewis be forthcoming (Patterson 1965, Derbyshire et aL <1929) this profile of steep lee slopes is typical of 1965, p. 3). The present investigation is essentially most of the glaciated areas of Tasmania. In the a reconnaissance one and, in an area such as this, Frenchmans Cap area the effect is emphasised in evidence of multiple glaciation is not quickly avai.l-· places such as the Clytemnaestra Ridge where the able, it being generally recognised as more difficult resistant quartzite and quartz-schist has been worn to obtain in areas of alpine .glaciation than in areas away on the lee sides to expose softer underlying formerly covered by continental ice-caps. Ice rocks while the quartzite remains to cap the ridge streams are more likely to erode evidence of pre and form the western slopes. Post-glacial pro vious glaciations because they are confined to the cesses have tended to preserve and in some cases one route of movement. Even where evidence emphasise this general asymmetry on the crests exists, it is hard to distinguish unless by obvious of the north-south ridges, the very tops of which differentiation of weathering because of a general take on a striking resemblance to the crests lack of differentiation in texture, lithology and of mobile sand dunes due to the erosion caused colour of deposits of different ages, arising from by the growth of heavy snow cornices along the this that glaciers of penultimate and earlier glaciations and Ahmad 19571 To the west the great gorges of extended beyond the well marked limits of the last the Fury, Mackintosh and Canning Rivers belie at the Vera, Livingston and Tahune gorges. All glaciation but the tills exposed on the Murchison drift belonging to these glaciers would then have Highway are unlikely to have been deposited by been redistributed such that if it existed at all it ice of local origin. To the north the picture is as would now be very difficult to distinguish. yet far from clear despite its importance to current Two possibilities remain: either previous glacia engineering projects in that area. To the south tions were much slighter than the last or insuffi and east similar problems leave current thought in cient snow accumulated to form a significant a state of flux. (Derbyshire 1963, p. 108; Jennings quantity of glacial ice. Gentilli 0961, pp. 478-479) and Ahmad 1957, p. 74). argued on climatological grounds for the existence The uncertainty in defining the maximum extent of glaciations prior to the last while admitting the of the Cap valley glaciers has been mentioned in lack of relics from it. Probably older glaciations previous sections. Accordingly Fig. 2 should not be were no more extensive than the last, with peri regarded as final. In the light of these difficulties, glacial processes predominating for much of the future consideration of the maximum glacial time. In this case the resulting landforms would advance here and in the rest of Tasmania should have been destroyed in the interglacial periods by take account of the following possibilities:-·-· fluvial erosion. This would mean that the latest glaciers formed and advanced upon a fluvial land (1) If the last glaciation attained an early, scape. comparatively brief and extensive maxi mum, the evidence for this might be Therefore, bearing the difficulties in mind, the obscured by the subsequent periglacial conclusion is reached that the Frenchmans Cap processes. area shows evidence of only one glaciation. Present opinion (Jennings and Banks 1958) correlates this (2) A distinction must be drawn between the glaciation with the Wurm or Wisconsin glacial mountain gla>Ciers and the Central phases of the northern hemisphere. The suggestion Plateau ice cap, which at one time had is made that previous glaciations may well have a potential accumulation area of some occurred despite the lack of evidence. 600 square miles or more. In contrast, Certainly the evidence in the area correlates the area above the firn line in the readily with the most recent glacial phase. Preser Frenchmans Cap area cannot, at any vation of moraines and other glacial landforms, time, have been greater than six square the generally undeveloped state of drift weathering, miles. the presence of illite in some deposits (Charles worth 1957, p. 226), the depth and newness of the (b) Post-maximum stage sequence of events lakes despite aetive infilling, all strongly suggest The various retreat stages of the valley glaciers this. The equally fresh evidence of glaciation in have been dealt with above. The glaciation of the ·the nearby Linda Valley (where radio-camon Cap area almost certainly i.nvolved a synchronous dating of fossil wood found in varves puts the valley glacier maximum stage that must have initial advance there at approximately 26500 years grown from, and must have retreated to, the high B.P. CGill 1956) l is ascribed to the latest glacial cirques. It was not possible to date the valley phase, the radio-carbon date corresponding with retreat stage. Perhaps future work in this or the beginning of the Wisconsin glaciation in North other parts of Tasmania will provide details of the America. fluctuations of the last glaciation. If the New Zealand evidence can be taken as an indicator of Problems of the last glaciation Southern Hemisphere climate changes in general, The systematic significance of the present study a sequence of three gla.cial advances might be lies in relating it to ( i) the general glacial history expected with the last retreat beginning some of Tasmania and (ii) factors affecting ice accumu 14000 years B.P. after the Kumara 3 and Poulter lation in alpine areas in order that the climate advance of the Otira glaciation of South Island (see changes responsible may be correctly interpreted. Suggate 1965, p. 84). To this end some major problems of more than immediate significance are discussed. These are Present work so far in the Frenchmans Cap area the extent of the last glaciation, the sequence of has yielded only one datable deposit. A radio events in that glaciation, and the dating of these carbon assay of wood and leaf fragments taken events. from rock flour CIRQUES, SNOWLINES, AND THE GLACIAL trough wall (Deception Ridge). Small cirques AGE CLIMATES formed above the west side of the glacier in the Among the methods used for quantifying Quater lee of the west trough wall (Clytemnaestra Ridge). nary climate change is that taking cirque levels as The Tahune g-lacier had reached its maximum indicators of glacial age snowlines and tempera extent at the Tahune Creek gorge and may have tures. Davies (in Gill 1962, p. 245) and Galloway discharged a small ice tongue over the diffluence ( 1965, Fig. 2) have spoken briefly, and in general beneath Sharlands Knob. Meanwhile, the terms, of the Tasmanian snowline while at the ooauuJLtl! and Sophie glaciers were cascading down same time mentioning the problems involved in on to the Ness Creek Plains. The Whitham glacier defining it. The difficulties are due to the east had overflowed its cirque and the east Philp cirques. west climatic contrast across Tasmania and the contained active ice at this stage. The valley con- complications inherent in assessing the inter strictions and difi:luence and the hanging relationships of temperature and precipitation snouts that the glaciers at changes (Galloway op. cit., p. 609). were unfavourable for the building of terminal and much of the It may be assumed that snowline levels change debris was carried constricted and with progressive deterioration and amelioration of fast·-fiowing meltwaters. were responsible for climate during glacial times. It is important when the valley trains of the Loddon Plains and Ness suggesting former snowline elevations and the Creek Flats. corollary ,climatic inferences to define the stage of glaciation to which it refers. Maximum stage The post-maximum stands are better marked snowlines might be deduced from climatic data in some valleys than in others. In the Vera valley derived using Ahlmann's curve 0948): evidence the glacier thinned at the point of greatest con of Jormer lower limits of solifluction and tree striction, so that retreat took place rapidly to this lines being used to help establish temperature point. Ice retreated from the diffluence passes and while modern climatic records of temperature and a major stillstand was recorded in the main Vera precipitation (ablation season length, &c.) may moraine. After a slight advance .that emplaced also be used. a push moraine on top of the main moraine, retreat with six halts ,took place until an that remained Cirque stage glaeiation snowline levels on the of the glacier was in the Barron and Marilyn other hand are orographic snowlines, and, as cirques. The former cirque contains a moraine mentioned above, this is strictly a function of testifying to this halt in the retreat. climate and lithology (page 126) . Wind and pre cipitation shadow effects can be seen in both The retreat stages of the Livingston glacier are individual mountain ranges and across the whole less obvious but it seems reasonable to postulate Island. Cirque glaciation, except as part of an a retreat from the maximum ext,ent to the lower advancing hemicycle, is essentially a late-glacial end of Lake Gertrude with a major halt across the or "post-glacial " event. Late-glacial 'cirque ice deposits separating Lakes Millicent and Magdalen. is very sensitive to climatic factors, and, at least as The retreat of the Tahune glacier is not marked investigations to date are able to ascertain, existed by any one moraine ridge although there is ample mainly in the high discrete cirques which are drift on the floor of .the valley. The final stages forming the subject of further investigation. saw the ice retreating into the Tahune, Jetty and Nancy cirques. The moraine at the mouth of the SUMMARY OF GLACIATION Tahune cirque contains a large proportion of pro The following sequence of events for the last talus from the north face of the Cap. The lack of glaciation is suggested. Climatic deterioration moraine in the high cirque is as much due to resulted in the accumulation of snow and the for abrasion having reached below the zone of open mation of snow patches in the depressions of a weathered joints as to a postulated brief late·· ftuvial landscape, probably between 30000 and 26000 glacial cirque phase. years B.P. The early accumulation hollows were The Sophie and Gwendolen glaciers retreated provided by larger solution features in dolomite into their fiat hanging troughs .from their steeper areas and by structural weaknesses elsewhere. lower valleys. The former, being smaller and there Cirque glaciers formed, ice continued to accumulate, fore more sensitive than the other, built a series and valley glaciers advanced to a maximum. All of low moraines near the trough mouth while the this time the highest peaks and ridges stood above Gwendolen glacier left a more uniform, though also the glacial ice as indicated by the evidence of pro ridged, till cover. The Sophie glacier retreated longed frost shattering. into the upper Cwm from which limited outwash In the Vera valley, glacial ice accumulating in was derived for the deposit of a small delta in the Marilyn and Barron cirques advanced down Lake Sophie immediately below Lal;::e Charlotte. valley and over the low dolomite and quartz-schist The Marilyn cirque glacier left a shallow protalus ridges where the Vera Plains now stand, and dis cover on the lip. The Whitham glacier retreated charged over five diffluence passes at the east end to its cirque and melted without leaving a moraine. of ,the plains and also down the lower Vera valley The absence of debris is explained by the lower to the Vera gorge, before which it was joined by headwall. the Pickaxe and Davern glaciers flowing down After the glacial maximum the East Philp cirques short, steep gradients from heads on the north contained only stagnant ice or neve and behaved side of the Nicoles Needle Ridge. generally as nivation cirques or snow patch hollows. At the maximum stage the Livingston glacier Protalus ramparts formed in the northern of these reached the gorge above the Franklin River junction cirques as well as in the neighbouring nivation possibly over-spilling the low section of the east hollows. JAMES A. PWI'ERSON 129 ACKNOWLEDGMENTS FiNUCANE, K. J. and BLAKE·, Ji'., 1933.-Country VV est Uoast R(Jad and the Jane River. 'J'a,s. Most of the field work for this paper was done Type'Writte·n Hep't. (Unpub.). GALLOWAY, R. W., l9G5.--Late Quaternary Climates in Australia. while the writer was an honours student in the J. of Gcol., 73, pp. G03-GJ 8. Department of Geography of the University of GENTILLI, J ., 1961.-Quaternary Clhnates of the Australian Tasmania. Region. N.Y. Acad. Sci., 95, Art. 1, pp. 465~501. (;ILL, E. D., 1956.---Hadio-carbon Dating for Glacial Varves in The writer wishes to thank the staff of the Tasrnania. Aust. J. of b'ci., vol. 19, no. 2~ _p. 80. Geography and Geology Departments of the Uni -~-----~---, 1DG2<-See Spry, A. H. and Banks, M. H ..• eds., 1!Hi2. versity of 'Tasmania for their assistance in both HARTSHORN, J. H., 1958.-Flowtill in Southeastern Massaehus~ the planning of the field work and the discussion setr.;. Geol. Soc. Arn. Bull., 69, PP. 4!77-48'2. of results. A particular debt of gratitude is owing Hu•rcHIN::-iON, G. ~~-. 1957.----A. Treatise of Limnology. Wiley5 New York, 2 vols., val. L to Dr. J. L. Davies of the Geography Department JI R.S.-10. PAPI:<)RS AND PROCEEDI'!'{GS OF THE ROYAL Socn.rr'i 01'' TASMANIA. VOLUMJ.J 100 PLATE I.-Aerial view of the Vera Retreat moraines. Diffiuenee pass middle-left, main Vera moraine with push moraine crest lower c.entre, recessional moraines on proxima] moraine edge. Hanging cirque containing Lake M.arilyn upper right. Lake Vera middle right. F.P.J:l(< PLATE 1!.·-~-Aedal vie"v .from 'We~t of Clytemnae~tra looking north-north,veftt, aeross the Sophie cb·que \'alley into the Upper Cwm. C.ap !' PLATE IlL-Abraded rock lip of Lake Whitham chque. Looking southv,rest. PLATE IV .-~Asymmetrical ridge and lee side excavation. View from the summit o.f Clyternnaestra looking north to the Cap summit. Note quartzites overlying darker coloured dolomite::>, schists, and phyllites of the Scotchfire Group. [Platt: H.---Geography Department! University of 'fasmania. All other photos by the author.] ,, '