denskjold has described one on the east coast of Graham Land. The greater size of the bergs on the Atlantic Ocean than on the Pacific side of Antarctica indicates the greater scale of the ice-sheet towards the Weddell Sea than towards the Ross Sea. Moreover, after the reports of the latest expeditions, it appears probable that the larger and more numerous bergs that occur to the south of the Atlantic and Indian Oceans are not wholly comparable in their formation and structure to those found in the Ross Sea, in the neighbourhood of the Ross Barrier.
The whole question of the Ross Barrier and a barrier described by Nordenskjold on the east coast of Graham Land, which he calls an "ice-terrace," is most interesting. Buchanan and Nordenskjold have pointed out that these barriers, or ice-terraces, are composed of névé, not glacier ice, and with the Graham Land Barrier this especially appears to be the case. Névé, however, precludes the idea of flow, and we have the definite record of Scott that Barne, on visiting a dépôt Scott had laid down, found that it had "moved on." "Thirteen and a half months," says Scott, "after the establishment of the dépôt, he measured its displacement, and found it to be 606 yards. And thus almost accidentally we obtained a very good indication of the movement of the
