Vanda. During the winter the pond is frozen com- 2001- E pletely to the bottom; during the 1969-1970 austral EL 180- summer, it was ice-free except for an ice remnant that 9 160- accounted for approximately 35 percent surface area. Cr

On occasion, when the sun was below the mountain 140 range to the south during the early hours of the day, a thin film of ice would form over its surface. For 3 days—January 20-22, 1970—Canopus Pond was sampled at 2- to 3-hour intervals. Biological oxy- 120- gen demand (BOD) was determined for each sample 2 100- immediately after collection by a standard (APHA) in chemical method. The variation of dissolved oxygen concentration with respect to time is shown in fig. 1. Although no distinct maxima were evident, marked minima occurred every 24 hours at approximately 1100 hours. CL 240 An aliquot of each water sample was retained for CL chemical analysis. The concentrations of chloride, 200

calcium, magnesium, and sodium ions have been de- 60 termined on these samples. The variation of the ions with respect to time is shown in fig. 2. There is a peri- odic fluctuation of concentration, with minima occurring at approximately 12-hour intervals, at 240 1000-1200 hours and 2200-2400 hours. The variation of ion concentrations in Canopus 200 Pond over such brief periods has tentatively been at- 160 tributed to the algal population (Prof. E. P. Odum, 2 12 1 12 1/20/70 1/21/70 /22/70 personal communication). The main source of water TIME (hours) is a small intermittent meltwater stream originating Figure 2. Variation of concentrations of calcium, magnesium, from an alpine glacier to the south of the pond. No sodium, and chloride ions with time, Canopus Pond, Wright Valley. apparent correlation exists between stream flow and concentration of the ions. During the period of inves- The advice, assistance, and encouragement of Dr. tigation the amount of discharge into the pond could Derry D. Koob of the Department of Wildlife Re- not account for dilution of the salts, nor could the rate sources, Utah State University, are gratefully of evaporation produce the maxima observed in fig. 2. acknowledged. The assistance of the U.S. Navys Ant- The following algae were identified from Canopus arctic Development Squadron Six was invaluable dur- Pond: Hantzschia amphioxys (Ehr.) Grun. var. maior ing the field study. Financial assistance was provided Grun.; Phormidium fragile (Menegh.) Gom.; Navi- by the National Science Foundation through grants cula muticopsis van Heurk (tentative) ; and Stauroneis GA-14427 and GA-14573. anceps Ehr. (tentative). Work is continuing on these samples. Glaciology and meteorology of Anvers Island: E subglacial surface of Marr Ice wZ C, Piedmont >. X 0 ARTHUR S. RUNDLE uJ Institute of Polar Studies 0 U) The Ohio State University U) Analysis of data obtained from Palmer Station be- 1/20/70 121/70 1/22/70 tween February 1965 and January 1968 is nearing TIME (hours) completion (Rundle, in preparation). Data tabula- tions were published by Rundle et al. (1968) and Run- Figure 1. Variation of dissolved oxygen concentration with time, Canopus Pond. dle and DeWitt (1968). ANTARCTIC JOURNAL 202

BOO

IM -600

BED ROCK

DISTANCE IN KIT,

G4 LINDA HI 42 G3 GI -4OO 1 413 H4 HT H6 -200 ICE SL

Figure 1 (below). Location map 15 10 BED ROCK 15 20 ol Anvers Island and key to sub- DISTANCE IN glacial profiles.

-600 -- MUlI21 K3 FLAG K4 K15 -4000 tr erssubaci profiles; M Ice Piedmont. ICE ICE

SL BED ROCK 1 RED ROCK L 200

DISTANCE IN Kt, 13 14 T5 TI -800 R C6 i

ICE H6 0

BED ROCK

DISTANCE INK

NI N2 N4 -600

ICE BISCOE P01-r- 200

_- BED ROCK SL

DISTANCE IN Km

In a study of mass balance, Rundle (1970) sug- gested that the Marr Ice Piedmont is in equilibrium, or possibly slightly positive. Further analyses, using ad- ditional data not included in Rundle (1970), suggest that the balance may have been slightly positive for a prolonged period (possibly over the past 150 to 200 years) after a period of marked recession. The piedmont is typical of the so-called "fringing glaciers" of the west coast of the in that the ice, as seen at the coastal cliffs, is remarka- bly free of rock debris. This leads to the facile conclu- sion that the piedmonts are essentially protective agents. Conversely, Holtedahl (1929) believed that the piedmonts are "strandflat glaciers" which are actively cutting planed surfaces at a level controlled by the sea.

203 Ice thickness, measured gravimetrically (Dewart, in monts of Trinity Peninsula also rested on preglaci4.1 press), ranges from 60 to 80 m at the coastal cliffs to surfaces. His calculated bedrock profile from Depot more than 600 in Profiles plotted from these Glacier is remarkably similar to that of the Marr Ice data (figs. 1 and 2) show that the piedmont rests on Piedmont. Dewar (1967) described planed surfaces at two low coastal platforms, one at approximately 50 m, three general levels on including the other at approximately 200 in In places platform of unknown height and origin beneath the these surfaces are deeply dissected by valleys, resulting Fuchs Ice Piedmont. In the north, the Pecten Con- in pronounced ice streaming at the surface. This is not glomerate (Andersson, 1906) rests on a wave-cut plat - a strandfiat, according to the classical description. form, presumably Late Miocene-Early Pliocene, at From considerations of the distribution of surface 220-250 in sea level. A similar and probably velocities and calculated basal shear stresses, the pres- contemporaneous pecten deposit at King George Ii - ent study concludes that there is a zonation of basal land stands at 45 m. There is a "remote possibility of in conditions that determines the erosive capability of a platform at 305 Darwin Island" (Adie, 1964 the piedmont. In the high interior, limited basal ero- p. 30). sion is thought to be occurring, in association with a Correlation across these features is probably not i - sliding mechanism similar to that of Weertman mediately feasible because of the superimposed effects (1957, 1964) ; this is discussed in some detail by Boul- of differential faulting and uplift, and postglacial is - ton (1970). This ice is ultimately channeled into the static recovery followed by eustatic submergence. "streaming" ice. Sliding velocities in the ice streams However, the evidence points to the possibility of a se- are high, exceeding 100 m per yr (over 75 percent of ries of platforms at 200 to 300 in This is in- the surface velocity) in places, suggesting the possibil- ferred to include the higher subglacial surface, at 2C 0 ity of melting beneath them. In this case, any basal m above sea level beneath the Marr Ice Piedmont. debris entering the ice streams is lost. If basal melting Whereas these observations do not support a pure y in the ice streams is an overestimate of conditions and glacial origin for the subglacial surfaces, they do not basal erosion is active beneath them, its debris load eliminate the obvious fact that glaciation has beeln (the visible evidence of erosion) is below sea level and still is a powerful agent in the geomorphological when it reaches the coastal cliffs. The remainder of history of the region. With regard to the marine origin the piedmont appears to be frozen to bedrock and of the surfaces, considerably more data are required, must be protective, because refreezing of the liquid and correlation across these features is needed. Above phase seems essential to the erosion process. all, the morphology of the subglacial surface--&f Ade- Consequently, Holtedahis suggestion that the sub- laide Island and elsewhere would be most interesting. glacial surface is of purely glacial origin is unsatisfac- The field investigations were supported by National tory, because the present behavior of the piedmont is Science Foundation grants GA-165 and GA-747 to not as he described. The explanation that the ice is The Ohio State University Research Foundation. wholly protective is not acceptable, because the evi- dence points to selective but active erosion. References The origin of the subglacial surfaces must lie, in Adie, R. J . 1964a. Sea-level changes in the Scotia Arc and part at least, with other agents, and it is suggested Graham Land. In: Antarctic Geology. Amsterdam, North- that they are preglacial and of initial marine origin, Holland. p. 27-32. 1964b. Geological history. In: Antarctic Re- later modified by glacial action. Former sea levels are Adie, R. J . search. London, Butterworths. p. 118-162. indicated by raised beaches, marine platforms, and Andersson, J . G. 1906. On the geology of Graham Land. terraces throughout the peninsula (e.g., Nichols, 1960, Uppsala Universitet. Mineralogisk-Geologisk Institut. Bul- 19-71. 19645 1966; Adie, 1964a, 1964b; Everett, 1971), and letin, 7: the regional occurrence of levels at about 6 m, 10 to Boulton, G. S. 1970. On the origin and transport of en- in glacial debris in Svalbard glaciers. Journal of Glaciology, 12 m, and 50 well established. It is believed 9(56) : 213-229. that the lower subglacial surface of the Marr Ice Dewar, G. J . 1967. Some aspects of the topography and Piedmont may be part of the 50 in This has glacierization of Adelaide Island. British Antarctic Survey. been severely modified by later glaciation followed by Bulletin, 11: 37-47. Dewart, G. In press. Gravimetric observations on Anvers eustatic submergence (Hooper, 1962, p. 5). Island and vicinity. In: Antarctic Research Series. There is only scant evidence to support the regional Everett, K. 1971. Observations on the glacial history Øf existence of higher platforms. Nichols (1966) sug- Livingston Island, South Shetland Islands, . Arctic, 24(1): 41-50. gested that a prolonged period of fluvial erosion pre- Holtedahl, 0. 1929. On the geology and physiography of ceded Pleistocene glaciation in the Antarctic Penin- soine antarctic and subantarctic islands. Scientific Results sula, and Marsh and Stubbs (1969) ascribed a pregla- of the Norwegian Antarctic Expeditions 1927-1928, cial marine origin to many of the "planed" surfaces at 172 p. Hooper, R. P. 1962. Petrology of Anvers Island and ad- 305 m and higher elevation, on the eastern side of the jacent islands. Falkland Islands Dependencies Survey. S i- peninsula. Koerner (1964) believed that the ice pied- entific Report, 34. 69 p.

204 ANTARCTIC JOURNAL Koerner, R. M. 1964. Glaciological observations in Trin- ity Peninsula and the islands of Prince Gustav Channel, Graham Land, 1958-60. British Antarctic Survey. Scien- tific Report, 42. 45 p. Marsh, A. F., and G. M. Stubbs. 1969. Physiography of the Flask Glacier—Joerg Peninsula Area, Graham Land. British Antarctic Survey. Bulletin, 19: 57.73. ichols, R. L. 1960. Geomorphology of Marguerite Bay area, Palmer Peninsula, Antarctica. Geological Society of America. Bulletin, 71(10): 1421-1450. 638 151 AXES 1964. Present status of antarctic glacial geol- ogy. In: Antarctic Geology. Amsterdam, North-Holland, N p. 123-137. 1966. Geomorphology of Antarctica. Antarctic Research Series, 8: 1-46. tundle, A. S., W. F. Ahrnsbrak, and C. C. Plummer. 1968. Glaciology and Meteorology of Anvers Island, 1: Surface Meteorological Data for Palmer Station, February 1- December 31, 1965. Ohio State University Research Foun- dation. 374 p. \\ /• and S. R. DeWitt. 1968. Glaciology and -f------152 AXES Meteorology of Anvers Island, 2: Surface Meteorological _- 200 AXES Data for Palmer Station, January 1-December 31, 1966. Ohio State University Research Foundation. 404 p. 1970. Snow accumulation and ice movement on the Anvers Island ice cap, Antarctica: study of mass balance. International Association for Scientific Hydrol- ogy. Publication, 86: 377-390. In preparation. Glaciology and Meteorology of Anvers Island, 3: Glaciology of the Marr Ice Piedmont. Ohio State University Research Foundation.

Weertman, J 1833 8 . 1957. On the sliding of glaciers. Journal 2151-548 200 AXES of Glaciology, 3(21): 33-38. 138 AXES 1964. The theory of glacier sliding. Journal Figure 1. C-axis fabric profile through the ice sheet at Byrd of Glaciology, 5(39): 287-303. Station. Data obtained from horizontally sectioned cores. Contour intervals are at 4, 3, 2, 1, and ½ percent for 100-rn profile; 5, 4, 3, 2, 1, and 1/2 percent for 638-rn profile; 10, 5, 4, 3, 2, and 1 percent for 1137-rn and 2151-54-rn profiles; 20, 15, 10, 5, 2, and Analysis of antarctic ice cores 1 percent for 1259-rn profile; and 5, 4, 3, 2, and 1 percent for 1833-rn profile. ANTHONY J. Gow U.S. Army Cold Regions Research and overriding effect of increasing temperatures in the ice Engineering Laboratory sheet rather than to any significant decrease in stress level. Detailed fabric analysis of a number of fine- Analysis of ice cores from the 2,164-m deep drill grained, cloudy bands that occur abundantly in the hçile at Byrd Station continues at the U.S. Army Cold 1,250 to 1,800 m zone indicates that these bands prob- Regions Research and Engineering Laboratory. Re- ably formed as a result of highly localized shear. The c€nt research has focused on petrofabric analyses of sense of shear as inferred from the symmetry of folded the ice and petrographic and mineralogical studies of bands is compatible with the directions of ice move- volcanic ash layers preserved in the cores. ment measured at the surface. The fact that fine- Petrographic analyses of the ice. A comprehensive grained cloudy bands are also present in the very investigation of the petrofabrics of the Byrd ice cores, coarse grained ice below 1,800 in strengthens entailing measurements of c-axis orientations of more the idea that these bands are actively involved with than 10,000 crystals, has established definitely the exis- shearing in the ice sheet. tence of a structurally stratified ice sheet in West Ant- Petrographic and mineralogical studies of volcanic arctica. As demonstrated in fig. 1, a slow but persist- ash layers preserved in cores. Visible debris bands in ent increase in c-axis orientation was observed between the ice cores have been identified positively as being of the surface and 1,137 m depth. By 1,250 m the struc- direct volcanic origin, i.e., deposited as ash on the sur- ture had transformed into a fine-grained mosaic of face of the ice sheet. These ash bands (fig. 2) are crystals with their basal planes oriented substantially composed predominantly of glass shards, but crystal parallel to the surface of the ice sheet. This fabric fragments and lithic chips (mainly of andesite) are (which persisted to 1,800 m depth) is attributed to also present. It was subsequently discovered that shar deformation. A rapid transformation to multi- cloudy bands (provisionally identified as shear bands ple-maxima fabrics below 1,800 m is ascribed to the on the basis of c-axis fabrics and related deforma- September—October 1971 205