OEiaiN OF THE DEAIEAGE SYSTEM. 
219 
of the terraces we noted the fact that in the southern portion of the lake 
basin only the lower Eocene was deposited, while in the northern portion 
around the Uintas the whole Eocene formation is present. Whence we 
infer that the final desiccation of the lake began in its southern or south- 
western portions, and that the lake shrank away very slowly towards the 
north, finally disappearing at the base of the Uintas at the close of Eocene 
time. 
We must also infer that upon the floor of this basin, as it emerged, a 
drainage system was laid out. Such a drainage system would necessarily 
conform to the slopes of the country then existing. Taking the supposition 
already made, that the uplift was somewhat greater upon the eastern than 
upon the western side of the province, the configuration of the principal 
drainage channels would be very much like that now existing. The trunk 
channel would flow southwestward and westward, while the tributaries 
would enter it on either hand very much as the larger and older tributaries 
now do. The affluents on the south side are the San Juan, the Little Colo- 
rado, and Cataract Creek, which seem to be due to just such an original 
surface. On the north side of the Colorado the arrangement of the tribu- 
taries also seems to conform to the assumption. On this side the later 
movements of the strata have been such that the prevailing courses of the 
streams are almost always against the dips. But when we restore these 
displacements and deduce from them as nearly as we may the original 
conformation of the country, the positions of the tributaries at once become 
natural and easy of explanation. 
The argument here adopted concerning the origin of the drainage sys- 
tem affords little scope for discussion. Rivers originated somehow. It 
seems almost a truism to say that they originated with the land itself, and 
that their courses were, in the first instance, determined by the slopes of 
the newly emerged land surface. No doubt there are many causes which 
may have changed the courses of rivers, and in the subsequent changes of 
position the original arrangement may have been lost and left no intelligible 
trace. On the other hand, there are certain conditions under which we may 
look for the highest degree of stability in the positions of drainage chan- 
nels, and when we find such conditions to have prevailed continuously the 
