TOPOGRAPHIC FEATURES OF THE DESERT BASINS. AY 
turning of the river toward the basin in 1905 and 1906, a condition 
which it cost millions to remedy. Had no attempt been made to 
return the river to its seaward bed, or had this attempt been unsuc- 
cessful, the basin would have filled until it overflowed into the Gulf 
or until the Colorado turned southward of its own accord. In either 
case desertion by the river would have left an inclosed sea to slowly 
evaporate as the Salton Sea is now doing. That this history was 
actually enacted in the recent past is indicated by a deeply cut old 
beach line surrounding the basin at about 40 feet above sea level and 
a series of lesser and lower strands marking stages of retreat. The 
similarity of conditions then and now is attested by the fact that this 
older series of strands can not be distinguished from the strands 
which have been formed by the retreat of the present Salton Sea. 
This interpretation of the recent history of the Salton Basin may 
require modification in detail. For instance, there has been some 
degree of post-Tertiary movement along the north side of the basin, 
and the exposed beds have been found to contain saline strata which 
exactly simulate beds deposited in continental inclosed lakes or 
playas. It is difficult to reconcile this with the hypothesis of long 
marine occupation.of the trough. In this and other directions Dr. 
Blake’s theory may need revision, but its essentials will probably 
stand. In any case, it is apparent that both the topography and the 
history of the Salton have been very different from that of the basins 
previously discussed. The major factor has been, not varying 
climate but a vagrant Colorado. This difference of history makes the 
usual criteria of little value. 
The size and nature of the drainage basin, its mutilation by stream 
decay, the probabilities of inflow or overflow, are here of little im- 
portance. Of course, salt accumulation is quite possible either by 
the evaporation of marine water, the assistance of the river, or the 
ordinary continental processes, but the problem is in no case one of 
topography and is therefore beyond the scope of the present paper. 
A word should perhaps be devoted to the delta of the Colorado.1 
This is a broad, alluvial plain, of no visible relief, and traversed by a 
network of bayous. The position occupied by the divide between 
culf and basin is entirely indeterminate, and there is no rainfall to 
developit. The lower channel of the Colorado is exceedingly variable 
and the delta is dotted with lakes and marshes, which are souvenirs 
of its presence. So far as known, all of these are essentially fresh, 
except some small ponds near the so-called Volcano Lake, and the 
salinity of these is due to recent and present fumarole activity. One 
of the lakes contains a considerable percentage of potash alum, which 
will doubtless be developed when transportation and political condi- 
1 For information concerning the delta Iam indebted mainly to the papersof Dr. D. T. MacDougalandto ~ 
personal communication from him. 
