of saline water, whereas very little of such movement would occur in the case of the barrier ice, and consequently the marks of stratification remain visible in the bergs which are calved from it. Except for a certain amount of glacier ice, which comes in from the feeders mentioned, the Ross Barrier is made up almost entirely of successive years' additions of snow and drift that fall upon it and accumulate in definite layers. The simile, therefore, that I have already given of the snow layer on a roof is all the more striking, only it is not the accumulation of snow of a single fall, not even of snow of a single year, but probably of snow that has fallen, say, during a thousand years.
It would not do to pass by Nordenskjold's important observations with regard to his "ice terrace" at Graham Land, and it is best to quote his own words (Antarctica, Dr. Otto Nordenskjold: London, 1905) as follows: "At our noonday rest I was nearly falling into a broad crevasse, but said nothing of the matter, in order not to make the others anxious. But all of a sudden the ice became more uneven, and at 5 p.m. our march came to a sudden and unexpected end in front of a canal-like crevasse, some 20 metres (65 feet) broad and almost as deep, which seemed to run in towards the land as far as the eye could reach. This
