NEW LIGHT ON ISOSTASY. 333 
be converted into a plateau between ‘2000 and 2500 feet higher 
than the adjacent portion of the plain. The conclusion is thus 
reached that the whole mountain mass above _the level of its 
base is in excess of the requirement for isostatic adjustment ; or, 
in other words, is sustained by the rigidity of the earth. Three 
stations in Yellowstone Park tell the same story as to the Rocky 
Mountains of Montana, and single stations on the Wasatch 
Plateau and the Appalachian Mountains indicate that those 
uplands are rigidly upheld. 
These results tend to show that the earth is able to bear on 
its surface greater loads than American geologists, myself 
included, have been disposed to admit. They indicate that 
unloading and loading through degradation and deposition can- 
not be the cause of the continued rising of mountain ridges with 
reference to adjacent valleys, but that, on the contrary, the ris- 
ing of mountain ridges, or orogenic corrugation, is directly 
opposed by gravity and is accomplished by independent forces 
in spite of gravitational resistance. 
While the new data thus indicate that the law of isostasy does 
not obtain in the case of single ridges of the size of a large 
mountain range, they agree with all other systems of gravity 
measurements in declaring the isostasy of the greater features of 
relief. The mode of reducing gravity measurements at different 
places so as to make them comparable depends on the theoretic 
conception of terrestrial rigidity, one method being followed 
when high rigidity is postulated and another when isostasy is 
postulated. Under the postulate of high rigidity it is assumed 
that all parts of the crust have the same density; under the 
postulate of isostasy each vertical element of the crust is 
assumed to have the same mass, density being inversely related 
to altitude of the surface. If either of these postulates were 
absolutely true the measurements, when reduced in accordance 
with it, would become identical, except for errors of observa- 
tion; and approximation to such identity is a measure of the 
degree of approximation of the corresponding postulate to the 
actual fact. Mr. Putnam and the writer independently applied 
