Crosby.] 
130 
[May 21, 
ness of the feldspathic detritus. Not only are the bowlders and 
smaller fragments of the granitic and other acid rocks still hard and 
undecayed, but even the minute grains of feldspar resulting from 
the complete disintegration of granite are usually quite bright and 
sound, rarely yielding a sensible reaction for water when strongly 
heated. Among the more basic rocks, the very abundant diorite 
debris is rarely weathered to any noticeable extent, but the dia- 
base is generally pretty well decayed. It is doubtful, however, 
if the rotting of the diabase can be regarded as wholly postglacial, 
for there are good grounds for believing that in preglacial times the 
dikes of diabase were more deeply decomposed than any other 
rocks in this region. A unique and very conclusive bit of evidence 
that the small proportion of clay in the till is } r et sufficient to ef- 
fectually exclude the meteoric agencies is afforded by the pyrite in 
the Somerville slates. In the quarries the bright cubic crystals are 
speedily tarnished and in a very few years are completely changed 
to ferric hydrate ; but precisely similar ciystals. imbedded in frag- 
ments of slate and also detached from the matrix, have been re- 
peatedly observed in the lower till of the East Boston drumlin in 
a perfectly fresh and untarnished condition. This means that they 
have been hermetically sealed in the hardpan, and if undisturbed 
would have remained chemically intact until superficial oxidation 
had slowly leavened the mass to that depth. The naturally imper- 
vious character of the clay has undoubtedly been greatly augmented 
by the compacting of the mass due to the enormous and long con- 
tinued pressure of the ice-sheet. 
In the upper or buff till the silicates, both acid and basic, are 
quite generally kaolinized, at least superficially, and the ferric hy- 
drate attending the access of free oxygen appears to reside chiefly 
in the clay, the coarser constituents, when freed from the clay, be- 
ing essentially similar in color to those derived from the gray till* 
It is a difficult matter, however, to separate perfectly the several 
constituents of the buff till, the ferruginous clay acting as a firm 
cement to bind them together. The smaller pebbles and grains of 
sand are thus caused to adhere strongly to the larger pebbles and to 
each other ; and considerable masses of conglomerate are sometimes 
formed in this way. Since it was impossible to resolve all these 
aggregates, the analyses of the buff till tend to show an excess of 
coarse detritus. No bright feldspar and no unoxidized pyrite have 
been observed in the buff till. The lower limit of the oxidation, 
