Crosby.] 
126 
May 21, 
forward by the ice, which will receive farther consideration in the 
concluding paragraphs. 
As we should naturally expect, the gravel is composed chiefly of 
fragments of the harder and more massive rocks of the region, 
quartzite, sandstone, quartz, granite, felsite, diorite, diabase, etc., 
except where the drumlin affording the sample lies immediately 
to the southward of a broad area of slate, and then the principle 
that the till is mainly of local origin asserts itself very distinctly, 
fragments of slate largely predominating in spite of their softness. 
This is particularly well seen in the East Boston and Convent Hill 
drumlins, which are in the lee of from three to four miles of slate 
underlying the Mystic and Malden Valleys. But even here it is 
very noticeable that while the slate forms a large proportion — 
three-fourths to nine-tenths — of the larger fragments, or those 
from one-half to two inches in length, the smaller fragments and 
grains consist chiefly of harder rocks from more distant sources, 
the slate rarely forming more than one-half and often less than one- 
fourth of the medium and fine gravel. This is an indication that 
below a certain fineness the slate is, to a large extent, completely 
pulverized — reduced to clay or mud. Among the larger stones 
and bowlders of the till, on the other hand, the slate is also rela- 
tively inconspicuous as we should expect in the case of such a soft, 
thin-bedded and finely-jointed rock. Even in East Boston, a care- 
ful examination and census, made under my direction by Mr. F. 
W. Atwood, of a large group of bowlders which were separated 
from the till during the grading down of the drumlin, show that 
only about one-third of the whole have been derived from the ad- 
jacent slate ; the actual percentages of the different kinds of rocks 
in 3412 bowlders being as follows: Diorite and diabase, 41.4; 
granite, 14.3 ; felsite, 1.1 ; quartzite, 8.1 ; slate, 33.7 ; mica schist 
and miscellaneous, 1.4. With the exception of the slate and an 
uncertain proportion of the diabase, all of these bowlders have 
been derived from the massive eruptive rocks and quartzites from 
four to ten or twelve miles distant; and an inspection of the geo- 
logical map makes it very clear that here, as must of course be true 
general^, it is the hardness and resistant character, more than the 
remoteness or size of the exposed area, of any rock that determines 
its occurrence in the form of true “hard heads” or travelled bowl- 
ders. The vitreous quartz, which is with quartzite such a promi- 
nent constituent of the medium and fine gravel, appears to have 
