during the melting of the Glacier. 423 
is, however, proved, and the general amount of pitch seaward 
educed cannot be far from the truth. 
For determining the actual amount of elevation along the 
lower Housatonic since the Champlain period no satisfactory 
data have been obtained. It is plain that the height of the ter- 
race at Birmingham is a measure chiefly of the vastness of the 
flood, and that the “necessary deductions” for reducing it 
to a measure of elevation are very large. The evidence with 
reference to the elevation derivable from the structure of 
the beds could not be studied for want of sections. The wide 
extent of the thirty-foot terrace-plain suggests that it may 
mark the Champlain water-level; but of this other proof is 
needed. The formation of the seashore terrace would not 
require an elevation of more than 12 or 15 feet. 
3. THe Lower Part orf THE CONNECTICUT RIVER VALLEY. 
its would have been in the condition of a lake one to four 
miles wide, and over part of the area, 100 to 150 feet deep. 
This Connecticut valley lake would have extended northward, 
harrowing and having shallow borders between Cromwell and 
Ellis, recently in charge of the Government Survey of the Con- 
necticut River, that the stream at low-water at Hartford (15 
miles north of Middletown) is on a level with mean low water 
Mm the ocean. Passing Hartford, it would have continued to 
Springfield and beyond; but it was probably in this part a 
, rapid stream, with shallow borders most of the way on 
either side of its main channel, and having part of the pitch 
of the present stream, about 384 feet in its course between 
