GLACIAL DRIFT. 217 
in a westerly direction; and this may indicate what sort of phenomena 
should be looked for, in the other westerly descending valleys mentioned, 
to satisfy us of the existence of tributary glaciers. No pains have been 
taken to observe this class of facts. Oliverian brook in Haverhill may be 
expected to show some signs of a local glacier descending the west flank 
of Moosilauke. On page 30, considerations are presented to show that 
unusually abundant gravel and sand deposits in Haverhill must have 
been derived from such a source as this local glacier would be. We have 
also a barrier-like ridge of till lying across the Oliverian valley suggestive 
of a former terminal moraine. There are many boulders of Bethlehem 
and porphyritic gneiss in the Oliverian valley, seemingly of eastern origin. 
If not from the Moosilauke pile, they must have been brought down by 
the main Connecticut Valley ice. The numerous marks between Haver- 
hill and Carroll may be regarded as made by one branch of the Con- 
necticut glacier. These are very abundant, and satisfactorily illustrate 
the local movement. Those down the Passumpsic valley are less com- 
mon and convincing. 
Above the Ammonoosuc district the Connecticut Valley movement 
seems to have been interrupted in consequence of the presence of the 
Gardner Mountain barrier. The markings in this part of the valley do 
not conform to the topography, and must therefore be regarded as made 
by the older movement. The map (PI. II) illustrates the nature of the 
ground over this space. Above Lancaster there are well marked signs 
of a local glacier, probably the last movement in the region, separate 
from and later than that just described. Its existence is manifested by 
moraines and not by striae, and may be classed with the Bethlehem gla- 
cier, to be described presently in the words of the late Prof. Agassiz. 
Midway between Stratford Hollow and North Stratford is a pile of 
granite boulders so conspicuous as to arrest the attention of the most 
careless, They occupy several acres of ground, and seem to form a ter- 
minal moraine. It is represented in Fig. 53. The exact location is indi- 
cated by the word Moraine on Pl. I; and a section crossing the valley is 
shown in Fig. 2. Mr. Clough’s stereoscopic view of the same is entitled 
“New Hampshire Cow Pasture.” The name indicates the use to which 
the field is put, while the abundance of boulders seems to have discouraged 
the owner from attempting to make a stone wall of the material. A 
