26 The Auicncan Geologist. January, 1903. 
terior region, adjoining- the northern ice-sheet, so that when 
the somewhat later knolly and hilly marginal moraines were 
amassed along its boundaries at pauses of its recession, a coarser 
modified drift formation of sand and gravel was spread along 
the river valleys by the stronger and faster flowing streams 
that then received the drainage of the ice melting. The valleys 
had already been reexcavated from their filling of loess. New 
floodplains of gravel, sand, and silt, brought by the waters from 
the ice melting and rains during the Wisconsin or moraine- 
forming stage, attained 'great thickness near the ice borders, 
but thinned more rapidly southward, so that their continua- 
tions in the great valleys are now covered, or have been mostly 
v.^orked over and redeposited, in the Postglacial alluvium. 
As fast as the ice-sheet withdrew and left the land bare, 
it was reelevated farther north to nearly the same extent of a 
few hundred feet as before in the upper Mississippi region. 
A wave of permanent uplift thus advanced toward the central 
area that was last occupied by the icefields. Not many thous- 
and years, probably, after the loess areas were uplifted, the 
basin of the vast Lake Agassiz, now represented by Lake 
Winnipeg, was reelevated, rising mostly 300 to 500 feet at 
the rate of about a half of a foot yearly, by a gradual move- 
ment advancing from south to north ; at nearly the same time, 
the coast of Maine and the Champlain and St. Lawrence valleys 
were uplifted to a similar amount ; and in the north central 
area of our continent, including Hudson bay, the uplift still 
appears to be in progress, averaging for some districts per- 
haps as high a rate as three to five feet or more in a century. 
The loess region may have been wholly raised to its present 
bight far more rapidly, like the Lake Agassiz basin, within a 
thousand years, permitting the erosion of the loess to take place 
very soon after its deposition. No longer interval than a few 
thousand years may therefore have separated the lowan stage 
of loess deposition, with its very slow river currents, from the 
Wisconsin stage of moraine acciunulation and attendant trains 
of gravel and sand washed down the valleys by rapid streams. 
The admirable investigation of the physical and chemical 
characters of our loess deposits by Chamberlin and Salisbury 
in the paper of their joint authorship. "The Driftless Area of 
the L"]5per Mississippi \'alley," published in the Sixth Annual 
