138 GLACIOLOGY 



doubt as the result of our observations that the sea ice in summer is corroded 

 by warm water from beneath, and that even in winter, where there is any obstacle 

 on the sea floor which causes an upward deflection of the deeper waters, there the 

 surface temperature of sea water is kept so high that sea ice cannot form. Under 

 these circumstances, there can be no doubt that differential erosion must be a 

 factor in determining the differential heights of the ice-cliff of the Ross Barrier, 

 but that it is probably not an important factor, as is suggested by the fact that 

 where the water of Ross Sea along the Barrier edge is particularly deep (over 

 400 fathoms), so that conditions would be unfavourable for upward currents, there 

 the Barrier ice is specially thin. There can be little doubt that the thinness of 

 the ice in the central section of the Barrier ice-cliff" is chiefly due to the fact that 

 this is just the portion which is farther removed from the sources of ice supply, 

 the great glaciers of the horst and the smaller glaciers of Carmen Land and 

 other lands bounding the Barrier on the east. 



SUMMARY 



The Ross Barrier is now, for the most part, a floating piedmont, the shrunken 

 remnant of a much vaster piedmont aground. Its surface shows both radial and 

 tangential undulations. The gently swelling radial thickenings of the Barrier 

 represent the prolongations, for hundreds of miles seaward from the shore-line, of the 

 great land glaciers, widely fanned out where they leave the land and plunge into the 

 sea. The tangential or concentric undulations, having mostly an approximate east 

 to west trend, represent pressure ridges chiefly normal to the paths of the main 

 glacier streams which constitute the radial ribs. Partly these undulations repre- 

 sent those gi-eat pressure ridges which start from the point where the land glaciers 

 meet the Barrier surface, and, as the whole Barrier slowly moves seawards, become 

 gradually transferred to the edge of the Barrier facing Ross Sea. Partly they 

 represent pressure ridges, due to the stemming action of rocky islands smothered 

 under the ice, like the one discovered by Amundsen to the south of Framheim. 

 Such low-lying areas of the Barrier as have a radial trend owe the relative thinness 

 of the ice, of which they are formed, (a) either to the glacier ice ribs becoming thin 

 through fanning at their ends and at their sides, whei*e they become welded 

 together along their planes of contact, or (b) to the fact that these adjacent 

 floating piedmont glaciers, failing to touch one another at their sides, never- 

 theless become linked up by old sea ice, which, being protected from erosion by the 

 glacier ribs or ice jetties on either side, accumulate as "bay" ice from year to 

 year, until it acquires a considerable thickness. This thickness is constantly 

 increased by additions of snow, whether derived by drifting from old falls, or 

 originating in snow new fallen in situ. This process appears to be taking place 

 at the present time at the Drygalski-Reeves Piedmont, which is such an important 



