GEOLOGY OF NEW HAMPSHIRE 61 
abnormally compact lower till under the ordinary ground 
moraine. 
The writer can find no reference to the drumlins earlier than 
his own descriptions of Prospect Hill* in Andover, Mass., in 1867. 
It represented something accumulated by ice, but not an ordi- 
nary moraine. These hills received much attention in the report, 
hundreds of them having been mapped in theatlas. Their longer 
diameters were found to coincide with the direction of the ice 
movement ; being generally southeasterly in Rockingham county, 
southerly west of Merrimack River, and west of south in the 
Connecticut Valley on the edge of Massachusetts. Mr. Warren 
Upham devoted himself to the exploration of these lenticular 
hills, and in searching for them beyond the limits of New Hamp- 
shire, became interested in the terminal moraines of Cape Cod 
and Long Island.’ 
Special attention was given to the modified drift in the report 
by Mr. Upham. The terraces of the Connecticut and Merrimack 
Valleys were carefully mapped and leveled, with the intention of 
testing the application of the marine or fluviatile theory of their 
origin. It would appear that the highest terraces and deltas of 
tributaries represent the remnants of the ancient flood plain where 
the river had its greatest volume. These remnants are quite 
uniformly nearly two hundred feet above the Connecticut River— 
whether at the state line on the south or at the mouth of the Pas- 
sumpsic. Very commonly a tributary increases the height of the 
flood plain forashort distance. If the terraces had been made in 
a series of lakes, or at successive heights of the ocean, they should 
have been arranged in a series of steps from the sea upwards. 
Eskers had been described in Maine, but it remained for Mr. 
Upham to bring them to light in all the principal New Hampshire 
valleys, especially in the Connecticut from Windsor, Vt. to Lyme, 
N. H., a distance of thirty miles. The main river cuts its way 
across this gravel ridge seven times, and it is quite concealed by 
terrace deposits in a part of its course. 
* Proc. Essex Inst. Nat. Hist., Vol. V., p. 159. 
? Geol. New Hampshire, Part III., p. 300. 
