( 83 ) 
cates a rapid stream flowing over rocky ledges, and heavily 
charged with grit. 
A third hypothesis, and one that appears to be the most 
reasonable, is that during the presence of the ice sheet, a 
sub-glacial, or perhaps in part a supra-glacial stream down 
the upper valley of the Bronx found its way out over this 
ridge and began to cut it down; being prevented issuing by 
the old channel because of the presence of the ice. The 
objections to this are the brevity of the time allowed by these 
conditions for excavating a gorge 60-70 feet deep. The 
hilltops about the gorge are glaciated, as, indeed, is the sur- 
face of the country very generally in this vicinity. A lobe 
of entirely stagnant ice in the old channel as a diverting 
cause is regarded as an almost too temporary affair. The 
terrace of cobble stones is assuredly connected with the great 
floods of the ice period in some way and with very copious 
and swift waters, as their size indicates. 
Finally, it may be suggested, as a fourth hypothesis, that 
the present channel has always been the drainage line of the 
Bronx, to which it has consistently adhered, while the west- 
erly depression has been caused by the small stream now oc- 
cupying it; and that the brook has excavated this valley ata 
little slower rate than the Bronx has its present one. But 
when one sees the size of the west depression and the insig- 
nificance of the present stream it is clearly impossible that 
such could have been the case. 
It will at once occur to all who are familiar with the prob- 
lems involved in the river drainage of the State of Connecti- 
cut* that both the Housatonic and the Connecticut rivers have 
left what appear to be their natural channels and have turned 
eastward through ridges of gneiss, but the diversion of the 
Bronx, with its 50 or 60 feet of gorge and its potholes, is not 
a phenomenon of the same magnitude with that of the Con- 
necticut, the bottom of whose gorge lies four huudred feet 
* See in this connection W. M. Davis, Topographic Development of the 
Triassic Formation of the Connecticut Valley, Amer. Jour. Scz., June, 1889, 
423. H. B. Kiimmel, Some Rivers of Connecticut, Journal of Geology, 
Vol. I., 371, 1893. 
