852 REVIEWS 
and the Archipelago north of the continent are included. The Lau. 
rentide and Cordilleran ice-sheets are differentiated and the debatable 
belt between them indicated. The Keewatin ice-sheet which has since 
been differentiated from the Laurentide glacier in part at least by 
Tyrrell is, of course, not separately represented. The recession of the 
ice-sheet and the courses of the ice movement in the immediate vicin- 
ity of Lake Agassiz are very fully set forth, as well as the drift 
deposits of the region. The succession of terminal moraines is amply 
delineated by text and maps. The moraines from the seventh or 
Dovre to the eleventh or Mesabi are regarded as contemporaneous 
with Lake Agassiz. 
With this very ample but needful introduction the history of Lake 
Agassiz is delineated. A distinction is drawn between the Great Basin 
lakes, Bonneville, Lahontan, and others, and true glacial lakes of the 
Agassiz type. The indubitable evidences of the existence of the lake 
in a well-cut outlet, eroded cliffs, beaches, deltas, and lacustrine 
deposits, are set forth in general terms at the outset and taken up in 
much detail afterwards. 
The contemporaneity of the great ice-sheet and the dependence of 
the lake upon the ice mass for its northern barrier is a central point 
of interest in the monograph. ‘The changes in the history of the lake 
are made dependent upon the shifting position of this ice barrier, 
upon the erosion of the outlet, and upon progressive changes in the 
earth’s surface. An attempt is made to measure the duration of Lake 
Agassiz by means of its beaches, its moraines, and correlated phe- 
nomena, with the result that the period is believed to have been short 
and the formation of the moraines very rapid. Some alternative inter- 
pretations by Chamberlin, under whose direction the work was 
prosecuted, are introduced at this point at the request of the author, 
the chief purport of which is to assign a series of rising as well as fall- 
ing stages to the shores of Lake Agassiz and to thereby make the 
moraines antedate the highest beach and to leave the time occupied 
in their formation undetermined. 
The beaches are divided into two classes, the one set being those 
connected with the southern outlet at Lake Traverse and the other 
set those connected with some undetermined outlet to the northward. 
Five distinct stages, represented by as many beaches or groups of 
beaches, belong to the first set and four to the second. One of the 
most important features of the monograph is the accurate determina- 
