622 REVIEWS 
resulting location of shore lines and channels of outlet, an account is 
given of the actual positions of the ice-dam within the district and of 
the resulting series of glacial lakes. A map of moraines includes not 
only data from various sources previously publisht, but important new 
material, especially for the Canadian peninsula, and is adjusted to the 
subject in hand by the indication of the theoretic continuations and 
connections of partly surveyed moraines. Other maps show the glacial 
lakes, outlined with detail and confidence not only on the land side, 
where their shores are still preserved as dry beaches and terraces, but 
also the ice side, where all actual vestige has necessarily disappeared. 
Lake Maumee, in the western part of the Erie basin, lasted during 
three oscillations of the ice front, growing larger with each shitting of 
the dam. At first its discharge was at Fort Wayne and thence down 
the Wabash River; afterwards part of its surplus escaped across the 
“thumb” of Michigan at Imlay, running westward to the Lake Michi- 
gan basin. In the later part of its life the broad upland of the penin- 
sula of Canada was a nunatak. Lake Whittlesey, succeeding Lake 
Maumee in the same basin, held place for a single morainic cycle. The 
Canadian upland, no longer a nunatak, formed part of its northeastern 
shore, separating two ice lobes. Its discharge crost the thumb of 
Michigan at Ubly to a lower glacial lake, Saginaw, and thence ran 
westward through the Pewamo channel across the lower peninsula of 
Michigan. Lake Warren, uniting Lakes Whittlesey and Saginaw, at 
the level of the latter, endured for four morainic cycles, being 
greatly modified in outline by the successive changes of the retaining 
dam and becoming eventually much larger than its predecessors. 
It is to be noted that the author did not delay the publication of 
his generalization until the ground had been wholly covered by obser- 
vation nor until all the observations made use of had been verified. 
While guarding against misapprehension by constantly drawing the line 
between fact and theory, he has freely used his working hypothesis for 
purposes of local interpolation. ‘To whatever extent his local gener- 
alizations are deductions from theory they will eventually, through 
the extension of observation be made to serve as tests of the theory. 
Taylor dwells on the peculiar significance of the Ubly channel as 
an evidence of the existence of glacial lakes and a glacial dam, and 
closes with a general argument (drawn out by criticisms of J. W. 
Spencer) in support of the fundamental theory that the ice sheet 
served as a dam for the retention of lakes. 
