66 ME. F. DEBENHAM ON A NEW [vol. lxXV,, 



either our figures are very wrong indeed, or that there is addition 

 to the sheet in the only other possible way, by freezing from below. 



When we come to examine the possibility of this taking place 

 through so great a thickness we are again met by a lamentable 

 lack of data. The question may be put in this form : — supposing 

 that the ordinary sea-ice were to remain attached to the land for 

 many seasons instead of breaking up and floating away, how thick 

 would it ultimately become ? The answer seems to depend entirely 

 on the situation of the ice with regard to sea-currents, as is shown 

 by the few data that are at our disposal. 



Off the end of capes, such as at Hut Point, Cape Armitage, and 

 Cape Evans, where the water is shallow and the strength of the 

 current considerable, the ice which had grown to a thickness of 

 6 or 7 feet during the winter was thawed through again by the 

 sea-water by the end of the summer. On the other hand, in a 

 protected bay like that south of Hut Point the sea-ice does con- 

 tinue to increase during the second year, if it stays in. In 1902 the 

 ice formed in this bay reached a thickness of 10 feet, and stayed in 

 for the summer. By the end of the next summer it had increased 

 to 15 feet, and there is every reason to suppose that if it had still 

 continued to stay in it would ultimately have frozen to the bottom, 

 although this spot is only half a mile from the shallow patch 

 off the cape where the ice is thawed through each summer from 

 below. 



The existence of floes of old bay-ice up to 30 feet thick, some- 

 times met with in the pack in Boss Sea, shows that conduction of 

 heat is considerable through that thickness of ice, even when deeply 

 covered with snow. The governing factors of freezing at these 

 depths would appear to be the conductivity of the sheet itself and 

 the amount of movement in the water underneath it. Comparing 

 the McMurdo Sheet with the Barrier we get these peculiar contrasts : 

 the conduction of the Barrier would be the minimum, on account 

 of its great thickness and the large amount of air included in it ; 

 that of the McMurdo Sheet would be the maximum, on account of 

 its comparative thinness, and because it is formed for the greater 

 part of clear ice. 



All that we really know of the movements of the water under 

 the Barrier is that there is a constant drift from east to west 

 along its face. In McMurdo Sound, on the other hand, although 

 there is a strong current coming round Cape Armitage and running 

 up the western side of Boss Island, it seems to be very local, and 

 there is nothing to show the existence of a current coming from 

 under the ice-sheet in the middle of the Sound. Icebergs and 

 pack-ice coming down the Sound under the influence of a northerly 

 wind invariably drift back very slowly on the western side and 

 very rapidly on the eastern. 



It seems therefore that the conditions for the increase of the 

 sheet from below are favourable enough to warrant our assumption 

 that this is the explanation of the existence of the sheet long after 

 it should have disappeared under the effects of surface thaw. 



