Correspondence. 395 
dams". My faith in the uniformity of nature is not less strong than 
that of Mr. Leverett, and for this very reason when he includes among 
"established facts" glacial dams, based upon evidence which is found to 
be elusive, and when he ignores phenomena which he says have received 
less attention than they merit, although they are of fundamental im- 
portance, one's reason compels him to halt before such a doctrine, .and 
to discredit the acceptance of an hypothesis against which such powerful 
facts appear. 
Another class of phenomena embraces the channels across divides, 
frequently characterized by gravel floors, and where such are found 
they have often been considered as evidences per se of the outlets of 
glacial lakes. Against this interpretation I have already pointed out 
that we find terraces of similar hight upon both the southern and 
northern sides of the plateaus, — or outside and inside the glacial dams. 
Furthermore phenomena exactly similar to the so-called outlets of 
glacial lakes are seen at low altitudes within a few degrees of the 
equator, as for example in the Tehuantepec isthmus in Mexico, an 
illustration of which may be seen in the /accompanying figure. Upon 
the Atlantic side there is an extensive gravel terrace corresponding to 
Fig. I. Nortlioi'ii i-nd of cliamicl or geological <!anal over the 'ri'liuantci'^'c divide. 
the gravel floor of the channel across the divide, which is an exact 
reproduction of the so-called glacial lake outlets of the noith. This 
geologi'Tal canal is less than a mile long and a hundred and fifty feet 
deep. Upon the Pacific side the descent is so rapid that the corre- 
sponding terrace-like features have been washed away. The character- 
istics of this channel over the divide are almost like those at Crawford 
