THE ROSS BARRIER 139 



key to the structure of the Ross Barrier, but at the same time, as much of it is 

 covered by moraines, it may be nearly all fanned-out glacier ice. 



The Ross Barrier is thus formed literally of " thick-ribbed " ice with transverse 

 pressure ridges. It may be compared to a shield formed of a wicker-work frame 

 covered with hide, in which the rods represent the glacier ice jetties, and the smaller 

 osiers the pressure ridges, while the drift and fallen snow represent the hide. 



The exact relation between the amount of ice that is being annually added to 

 the Barrier from its great marginal glaciers, as well as from the drift snow and new- 

 fallen snow on the Barrier surface, to what is lost by ablation, is not yet known. 

 Near Minna Bluff the surface of the Barrier was added to at the rate of about 15 

 inches (380 millimetres) of dense snow, almost neve, equal to 7\ inches (190 

 millimetres) of rain annually. 



At the same time, the evidence already given, in the chapter entitled Meteoro- 

 logical Notes, makes it clear that on the whole, in the western portion at any rate, 

 of Ross Island ablation is in excess of precij^itation, as is suggested in particular by 

 the evidence of Sunk Lake, where the surface of the lake ice, only about 60 yards 

 (55 metres) distant from the sea, is actually 18 feet (5*48 metres) below sea-level. 



It is also to be noted that no information is at present available as to the rate at 

 which the base of the Barrier ice sheet is being melted in the water of Ross Sea. 



That the amount of ice annually lost by melting must be very considerable is 

 obvious from the following consideration : — Evidence shows that the sea ice of Ross 

 Sea, which in a single season attains a thickness of about 7 feet (2' 13 metres), is about 

 half melted away during the summer, the melting taking place from below upwards : 

 thus late in January the sea ice, formed during the preceding winter, is found to be 

 full of corrosion hollows. Some of these hollows undoubtedly have been formed by 

 the seals, which have bitten away the ice to form their breathing holes. Abandoned 

 later by the seals, these old breathing holes become roofed over with young ice. But 

 the number of these is negligible as compared with the millions of corrosion hollows 

 which honeycomb the sea ice of Ross Sea late in summer. It would aj^pear that 

 near lat. 77° S., at least 1 metre of ice is annually lost by melting. 



Evidence, already quoted earlier (in last part of chapter on Physiography, deal- 

 ing with Ross Sea), shows that some of the deeper water of Ross Sea must be warm 

 enough to melt sea ice, and therefore, of course, land ice also. This suggests the 

 important consideration that as the result of relatively warm currents flowing south- 

 wards under the eastern and central parts of the Ross Barrier melting may be going 

 on there continually during the winter as well as the summer, and an appreciable 

 thickness of ice is jjrobably annually removed from the base of the Barrier through 

 this cause. It would be of great interest and importance to ascertain what this 

 amount is. An approximation might be made by experimenting with blocks of ice, 

 which might be used as " control sjoecimens," at such sheltered spots on the Barrier 

 edge as Western Inlet and Bay of Whales. 



