602 THE FRESH-WATER LOCHS OF SCOTLAND 
of December or the beginning of January. It remains bound for 
a period of four to four and a half months, the ice-cover being 
sometimes 9 J feet thick. Wide cracks in the ice appear at intervals, 
and on the broken sheets coming together again the ice is piled up in 
heaps, called toros."" These crevasses, which have a breadth of from 
3 to 6 feet, or more, are sometimes about half a mile long, and form 
a serious impediment to communication. Sledge traffic lasts for three 
months, but at the end of April the ice melts near the shore and 
softens. The breaking of the ice-surface, as in the Alpine glaciers, is 
accompanied by a loud crash, recalling an explosion, followed by a 
rumbling noise. The crack is instantly filled with water to the 
level of the ice-surface, forming a kind of river. In eight to four- 
teen days it freezes again, and a new crack appears at another place. 
The ice melts slowly, the process lasting nearly two months. 
The fauna of Lake Baikal bears a close resemblance to the marine 
fauna,^ but on account of the great distance of the glacial Arctic Ocean 
and of the Pacific Ocean, it is difficult to suppose that the fauna of 
the lake had any connection whatever with the oceanic fauna, and 
besides its waters are quite fresh. The German geographer, Peschel,^ 
holds that Lake Baikal was in time past a gulf of the glacial Arctic 
Ocean, which in the Tertiary epoch probably covered the whole of 
Eastern Siberia. The German geologist, Neumayer^ sustains this 
opinion, according to which Lake Baikal is a relict lake. Chersky 
refutes this, and says that the Arctic Ocean did not extend so far. 
Hoernes, on the other hand, points out the resemblance of the 
molluscs of the family Hydrobiidse to the fossil shells supposed to 
have been derived from the great inland sea, Sarmate, which stretched 
from Graz and Kra'ma to the mountains of Thian Shan, and 
covered almost all Central Russia in Miocene and Tertiary times.* A 
third view is that of Androussoff, according to which the great depth 
of Lake Baikal, and the similarity of its external conditions to 
those of the sea, might enable the fresh- water Crustacea to form 
original species resembling marine forms. These are, of course, only 
hypotheses, but the fauna of the lake is very interesting ^ from the 
theoretical point of view, and merits further study. 
A most interesting and little-known fish, characteristic of the 
1 See Schokalsky and Sclimidt, op. cit., p. 50. 
2 Ibid. 3 jii^^ 
4 BogdanoYich {JVorks of the Tibet Expeditio7i, ] 889-90, vol. iii. p. 60) says that 
the fossils from the mountains near Kashgar, described by Stoliczka as Triassic 
and taken as confirming the supposition of tlie existence of the Sarmate Sea in 
Mesozoic times, were in reality Devonian. There appears to be no trace of either 
a Mesozoic or a Tertiary sea in that area, and it may be assumed that Central 
Russia has not been under the sea since Palaeozoic times. 
•5 See p. 359. 
