sae BULLETIN 54, U. S. DEPARTMENT OF AGRICULTURE. 
regards them as due to wind erosion. The similar though smaller 
wind-eroded hollows of the Estancia Basin, New Mexico, were noted 
on page 49, and the writer has seen analogous depressions in the 
Alkali and Warner Basins, Oreg. Both volcanic and eolian basins 
are likely always to be too small to have importance for the present 
study. This is certainly true of all known American examples. 
THE POSSIBILITIES OF POTASH. 
In the preceding pages there are named specifically nearly 200 
inclosed basins. Some of these are so obviously unimportant as to 
require no further mention. One hundred and twenty-six, which are 
somewhat more important, are given in Table I, with the area and a 
brief description of each, the arrangement being the same as that 
followed in the text. It is certain that any basins of possible value 
for potash will be included in this list, and it is just as certain that 
many that are included will have no possible value. Many of the 
latter are easily eliminated. First, it is obvious that no accumulation 
is to be expected in a basin which has recently overflowed either into 
the sea or into another basin. Applying this to the list of Table I, it 
is possible to eliminate from further consideration 62 basins, of which 
16 were once tributary to Lahontan, 9 to Bonneville, 11 to the 
Amargosa and the Mojave, 7 to miscellaneous inclosed basins, and 19 
more or less directly to the sea. These totals include 10 basins the 
previous drainage of which is not absolutely certain, though ex- 
tremely probable. These are the Clover Basin, the Goshute-Steptoe 
Valley, the Murray Valley, the Ralston Valley, Stonewall Flat, Sum- 
mer Lake, Long Valley (northwestern Nevada), Duck Flat, the Big 
Smoky Valley, and the Smiths Creek Valley. The nature of the 
doubt in each case can be ascertained by reference to the preceding 
chapter. 
A second general elimination can be made on the ground of small 
area. It is difficult to set exactly the limiting area which a basin 
must have in order to be promising, but it seems probable that basins 
which cover less than 500 square miles may safely be disregarded. 
Their potash deposits, if existent, would doubtless be small, and de- 
tailed prospecting would scarcely be warranted at least until larger 
basins have been explored. Applying this criterion to the remaining 
basins of Table I, we eliminate 10 more, Alkali Lake, Garfield, Teels, 
Huntoon, Goldfield, Sheep Range, Willard, Granite Mountains, Owl, 
and Encino. It is possible also to eliminate 12 others which were 
very probably drained, but which, in any event, are smaller than the 
upper limit which we have set. These are Warm Springs, Allan 
Springs, Mesquite (part of the Pahrump), Acme, Luning, Mina, 
Monte Cristo, Kawich, Yucca, Aurora, Deep Springs, and Pinos Wells. 
The conclusion that these small basins lack practical value does not 
