Jieciew of Recent, Geological Literature. 815 
sea, and to Eschscholtz l)ay and the Kovvak and ^'ukon rivers in 
Alaska. These fossil accumulations of ice, covered by a thin soil in 
which occur shells of Cyclas, Valvata, etc., wood of aider, willow and 
dwarf birch, and abundant bones and ivory tusks of the mammoth are 
especially extensive in the New Siberia islands. The great Lyakhoff 
island, having an area of about 2,000 square kilometers or 700 sqviare 
miles, is wholly thus underlain by ice, excepting four granite peaks, and 
the fossil ice has a similar development in the more northern islands of 
this group. 
Baron Toll ascribes the thin overlying clay and sand to the action of 
wind and water as eolian and lacustrine deposits; but it seems worthy 
of inquiry whether in the New Siberia islands and many other places 
they may not instead be derived mostly from englacial drift which be- 
came superglacial by ablation of formerly higher strata of ice, as on the 
borders of the Malaspina ice-sheet south of Mt. St. Elias. In some lo- 
calities, as along valleys and avenues of drainage. Baron Toll considers 
the underground ice as remnants of frozen river waters during a former 
epoch of greater severity of cold, and there the overlying soil may well 
be of fiuviatile origin: but on these islands he finds, by the granular 
structure of the ice, that it is derived from snowfall, being a remnant 
of a previous ice-sheet. Several views show marginal ice-cliffs on the 
Lyakhoff island having a vertical hight of al)out 60 feet, capjjed by two 
or three feet of soil. Likewise the description given by Dall for the ice- 
cliffs of Eschscholtz bay in Alaska, compaj-ing the texture of the ice to 
compacted hail, proves, as Baron Toll remarks and as was noted in the 
last April Am. Geologist (vol. xv. p. 258), that the ice there also is a 
remnant of an ice- sheet. In these places probably the overlying soil 
was englacial drift m a formerly luuch thicker sheet of land ice, similar 
to the Pleistocene ice-sheets of North America and Europe, though of 
smaller extent. 
The mammoth remains are never in the ice, but in frozen mud and 
•sand beds distinct from the fossil ice masses and often overlying them. 
During a time closely following the Glacial period, a warmer climate 
than that of the present day ijrevailed in Siberia and Alaska, enabling 
shrubs to grow two hundred miles north of their present limits, while 
herds of the mammoth and woolly rhinoceros, both of which have since 
become extinct, ranged north to the Arctic sea. The mild postglacial 
climate may have been due to a depression of the area about Bering 
strait, permitting a strong warm current from the Pacific to pass north 
through this strait. Its width now is 28 miles, with a nearly unifortn 
depth of 24 to 28 fathoms, and the i^resent currents vary in direction 
with stages of the tides. A moderate subsidence, which is indicated by 
the raised beaches of this region, similar to the subsidence of the drift- 
covered parts of N(jrth America and Europe during the Late Glacial or 
Champlain epoch, and probably contemporaneous with that epoch, 
seems therefore, in the ojjinion of the reviewer, to be the best explana- 
tion of the Siberian and Alaskan fossil remains of ice-.sheets and of ex- 
tinct mammals. When the ensuing re-elevation, .shutting away the 
