112 GLACIOLOGY 



Cape Royds is very probable, for in summer, when the sea ice and ice-foot had been 

 swept away, one could see huge boulders of the granite submerged at some distance 

 from the shore, their light colour making them easily distinguishable from the 

 surrounding kenyte. 



The moraine terraces above Blue Lake appear to be made up principally of 

 angular fragments of a dark, fine-grained volcanic rock, with phenocrysts of augite 

 and olivine. Among these fragments were a good many hexagonal prisms of the 

 same rock remarkably regular, some of them from 2 to 3 feet long. A number 

 of blocks of tuff were also present. The commonest and largest blocks were a 

 yellowish, fine-grained friable tuff. 



Along the eastern shore of Clear Lake it was noticeable that the only prominent 

 erratics, other than basalt and kenyte, were a yellowish-green trachyte, which 

 appears to have been the oldest of the Erebus series of eruptions. Fragments of 

 diabase or quartz-dolerite, from the huge sills of tliat rock on the mainland, were 

 of frequent occurrence. 



In regard to the moraines left by the glaciers from Erebus when, subsequent 

 to the recession from McMurdo Sound of the former Ross Barrier, they were able 

 to over-ride the ground which it formerly occupied, we observed in the southern 

 portion of Blue Lake seven mounds of kenyte rubble. These we called eskers. 

 The general shape and distribution is shown on Fig. 45. 



The eskers appear to be composed of fine kenyte rubble, and may have been 

 deposited by sub-glacial streams from an old glacier formerly descending from 

 Mount Erebus into the southern basin of the Blue Lake.* 



Passing northward from Cajje Royds we reach Horseshoe Bay. Here, for the 

 first time since leaving Cape Barne, the volcanic series of Mount Erebus disappears 

 once more under a permanent cover of glacier ice. The latter is the piedmont 

 glacier of Horseshoe Bay. This glacier forms an ice-cliff on the eastern portion of 

 Horseshoe Bay from 50 to 60 feet high. This is the terminal face of an almost 

 inert mass of ice fed by the nh'e fields of the north-western slopes of Erebus. That 

 the ice is not in active movement is proved by the general absence of crevasses, of 

 which only small examples were observed. One of these had its sides fi'inged with 

 beautiful feathery ice crystals, which had evidently grown from the moisture in the 

 com^Daratively warm air coming up from the bottom of the crevasse. This moisture 

 became deposited on projecting points of the crevasse walls. 



Another phase of growth of these secondary ice crystals produced hexagonal 

 plates, sometimes as much as 7 inches in diameter, and comparable with those 

 found on Murray's dredging line. The latter crystals are illustrated in Fig. 1 of 

 Plate XXVI. Fig. 2 is a view across Horseshoe Bay looking towards Cape Bird. 



This piedmont glacier appears to be continuous as far as to Cape Bird. Near 



* We were unable to ascertain the natuie of the foundation on which tiiey rested. It may have 

 been either rock or ice. 



