344 ANNUAL REPORT SMITHSONIAN INSTITUTION, 1909. 
a great part of the Norwegian continental plateau. Certainly the 
whole of Barents Sea must have been covered with glacial sheets of 
this type. Although these are hypotheses, the observations of the 
Discovery authorize the conclusions; it is this which assures them a 
high general bearing. 
Like the other glaciers of this part of the antarctic world, the 
Ross barrier is in a retrograde condition. It has receded an average 
of 24 kilometers from the positions which were fixed by James Ross, 
but the retreat is in places equal to 35 and even to 50 kilometers. 
If one sums up the annual advances of the glacial masses during 
sixty years, an estimate is reached of the enormous surface of. ice- 
bergs which must have been detached from the barrier during this 
lapse of time. Captain Scott thinks that the barrier must have oc- 
cupied positions much mote northerly, that the retreat has been very 
rapid, and it is by this ancient extension that can be explained the 
many remnants of glacial sheets still adhering to certain points of 
“Victoria Land. Thus Lady Newnes Bay is filled with one of these 
remnants which the existing conditions of glaciation do not explain. 
The surface of it forms long undulations and its mass is probably 
floating. Of the same class are perhaps the barriers Drygalski 
(75° 30’ south latitude) and Nordenskj6ld (76° 30’ south latitude), 
of which it is not known whether they are glacial lobes or the frag- 
ments of ancient barriers. 
The ancient barrier then probably advanced to the height of Cape 
Adare. Captain Scott thinks that at first it rested on the bottom 
of Ross Sea and aided in leveling it. Then, the quantity of ice 
diminishing, it probably became floating and commenced to break 
up and retreat rapidly. It is not, then, beyond cur concern to know 
the composition of the bottoms of the Ross Sea. It has been de- 
termined that on the site of the former front observed by Ross the 
bottom is formed of a yellow, tenacious, and consistent clay contain- 
ing tests of foraminifera, frustules of diatoms, and spicules of 
sponges. There is likewise clay 10 degrees farther north near the 
Balleny Isles. The bottom of the Ross Sea is composed of mud re- 
sulting from the pulverization of rocks by the great glaciers of the 
south Victoria Land. 
Face to face with the grand development, in the main, of the gla- 
cial phenomena of terrestrial origin one is surprised at the slight® 
power and duration of the marine ice. This trait fixes the physi- 
ognomy of Victoria Land. The fields of ice may reach great hori- 
zontal dimensions, but their thickness in Ross Sea never exceeds 
6 English feet (1.80 meters). The hummocks never exceed 3 feet in 
MacMurdo Sound, and Mr. Ferrar asserts that the direct increase of 
the ice can scarcely exceed 8 feet, since it breaks up, he thinks, each 
summer. The cause of this mediocrity of the sea ice is thought 
to have been doubtless the temperature of the water of Ross Sea, 
