286 CAINOZOIC PAL.EOGEOGRAPHY 



immense mass of ice moving powerfully in a direction from south-west to north-east. 

 This must have entirely filled the whole of the Orleans Channel. Some idea of the 

 thickness of this ice stream may be formed from the fact that while the height of 

 the adjacent island is at least 200 metres, soundings were obtained near by in the 

 channel of 620 metres, consequently this ice stream must have been at least 800 

 metres thick. 



The same author * also states in I'egard to the maximum glaciation of Graham 

 Land that in bygone time the glaciation of these tracts was much more extensive 

 than to-day. For example, he states that in regard to the past glaciation of Moose 

 Island, "A single glance at our maps (Plates IV. and V.) will suffice to show that 

 such a direction of the movement of the ice was here possible only in case the ice 

 stream filled Gerlache Channel in all its breadth, and was forced to the north-east 

 by the pressure against the island range that forms the north-west side of the 

 channel" (p. 55). He adds, " Also in Hope Bay I found very clear evidence of the 

 earlier maximum glaciation. This is in the form of a rocky hill, 100 metres high, once 

 over-ridden by the adjacent valley glacier, the summit of which is now only 20 metres 

 above sea-level." He also notes, " On the southernmost point surveyed by our 

 expedition the Borchgrevink Nunatak, in 66° S. lat., Nordenskjold found, by means 

 of erratic boulders, that the island ice once rose 300 metres higher on the sides of 

 the nunatak than it does to-day." Thus in Recent, or at least Post-Pliocene geolo- 

 gical time, a vertical decrease in altitude of the ice has been proved to the amount 

 of 300 metres in King Oscar Land, and 200 metres in the neighbourhood of the 

 Orleans Channel. E. Gourdon, Geologist to Dr. Charcot's Expedition, records I 

 "Les moraines que j'ai rencontrees dans le chenal de Roosen, tout au pied du pic 

 Jabet que dans I'ile Doumer, temoigment egalement d'une ativite que ne correspond 

 pas k celle des glaciers actuels, et les blocs erratiques qu'elles renferment indiquent 

 nettement une extension plus grande." In the neighbourhood of Kaiser Wilhelm II. 

 Land, Drygalski has shown that erratic blocks cover the slopes of the volcano of 

 Gaussberg, built up of leucite basalts, to a height of 370 metres above the sea and 

 some 350 metres above the level of the surrounding land ice. Thus near Gaussberg 

 the inland ice was formerly fully 350 metres thicker than it is at present. J 

 These erratics of biotite gneiss, gneissic quartz felspar hornblende rock, horn- 

 blende biotite granite, all suggest the existence of continental rocks inland from 

 Gaussberg. 



Similar evidence of recent retreat of the ice is everywhere noticeable in South 



* Bulletin Geol. Instit. Upsala, vol. iii., " On the Geology of Graham Land," by J. Gunnar 

 Andersson, pp. 53-57. 



t Exp&lition Antarctique Frangaise (1903-5), G^ogiaphie Physique, Glaciologie, Pdtrographie, 

 par E. Gourdon, p. 117. 



J Dr. E. Philippi, Ueber die Landeis-Beobachtungen der letzten funf Sudpolar Expeditionen, 

 p. 7. Deutscher Sudpolar Expeditionen, 1901-3, Bd. ii., Kartographie, Geologie, heft 1, p. 51, 

 pp. 59-60. 



