1890 .] 
133 
[Crosby. 
probably still somewhat alkaline and imperfectly hydrated. On 
the other hand, it is, of course, possible that the process of kao- 
linization was, in a large degree, completed during the mechanical 
breaking up and transportation of the half -decayed rocks. 
But the strongest argument for the efficiency of glacial erosion 
remains to be stated. It is found in the rock-flour, the most abun- 
dant constituent of the till. While all the gravel and sand, and 
even the bowlders in the till might, conceivably, be regarded as 
simply residuary materials somewffiat modified by glacial movement 
and friction, it seems impossible to account for the rock-flour in 
this way. Detrital materials such as are here considered may, in 
general, be formed in three distinct ways : (1) subaerial decay 
(chiefly chemical) ; (2) aqueous erosion (fluvial and marine) ; (3) 
glacial erosion. To which of these three causes should the rock- 
flour of the till be assigned as an effect ? Careful examination of 
typical examples of residuary earths shows, what might have been 
foreseen, that they afford a true rock-flour only in the exceptional 
case where the original rock contains a large proportion of silica in 
a fine state of division. Although the bed of a rapid stream and 
a shingle beach are geological mills in which, however hard the 
rocks, a portion of the grist is ground exceedingly fine, they are 
quite out of the question as general explanations of any part of 
the till. We are thus limited to glacial erosion in the search for 
an adequate source of the rock-flour. And a multitude of observ- 
ers in every region of recent glaciers are agreed that while the 
ice streams produce detritus of all kinds, from the coarsest 
to the finest, they are peculiarly competent to form the im- 
palpable mud which, giving rise to the turbidity of glacial streams, 
is easily mistaken for clay, but is found on proper examination to 
be chiefly rock-flour or the finest mechanical detritus. A glance at 
a strongly scored and polished (glaciated) rock surface would seem 
sufficient to satisfy any one that the material must have been re- 
moved mainly in extremely small particles. The rubbing and 
grinding of the ice on the northern or stoss side of the ledges 
produces chiefly rock-flour and sand ; while on the southern or lee- 
crag side it breaks and rends (quarries) the rocks, forming chiefly 
bowlders and smaller fragments or gravel. The stoss slopes, being 
the gentlest and broadest, afford the most detritus, and hence the 
predominance of the rock-flour. 
Although the considerations here presented tend to emphasize 
