J. D. Dana—The overflows of the flooded Connecticut. 503 
at to Mettawampe (Mt. Toby) in Sunderland and Sugarloaf in 
e ide ye 
Connecticut Valley, forming plate III, of the same volume. A 
terrace-plain, numbered 2, is made to spread southward from 
Northampton, or rather from Greenfield more than twenty miles 
map, however, presents a suggestion of the author’s mind rather 
than a fact at the time established: for terrace No. 2 of North- 
ampton, the highest mentioned for that place, is according to the 
text, only 97 feet above low water in the river instead of 134 feet, 
which makes it nearly 40 feet too low to pass over the divide to 
Westfield.* 
he makes the terraces a result of river erosion. e terraces of 
the Connecticut Valley, in the view 1 have presented, are wholly 
of fluvial origin, and the highest terraces of the various’ valleys, 
whether over the lower or higher parts of the country, are ap- 
proximately measures of the flood-level in each. 
The proof with regard to these westward overflows is derived 
from the terraces of stratified drift over the regions. a 
1. Overflow at the Meriden divide-—On this overflow it is 
necessary here only to repeat in brief the facts from page 425 
* i in his Surface 
en eign sasged oa uy So siping lowest as No. 1 ye 
are vos. 1, 2, 3, re vely 32, 46 and 92 feet in height, the last 
206 feet above the Le as pa er ae Se Ne Yk Orne A 
yet this No. 3 has a t color on colored Mihaetat 
high led ae 
levels over try—which I have been to Bs 
mostly terrace-formations at high levels where there were water-courses during 
the era of the melting glacier, if not so no 
