288 
GILBERT. 
found that the details of crater forms accorded so perfectly 
with my theoretic conception of the results of impact and 
so imperfectly with the results of volcanic action that further 
consideration of volcanic possibilities was unprofitable. Brief 
examination of other theories, as they were discovered one 
by one in the literature of the subjqct, satisfied me that they 
could not compete with the impact theory as interpreters of 
crater form, and thus I was led to devote myself to the 
development of the impact theory. As one after another 
the obstacles in its path were found to be movable, and as 
one after another the obscure phenomena of the moon were 
found to be illuminated by its light, it gradually ceased to 
be viewed as a tentative explanation and was adopted as 
the real explanation. The tendency toward advocacy thus 
sprang from conviction, but it has been further promoted 
by the consciousness that there are many in my audience 
who do not share with the student of geophysics his concep¬ 
tion of the plasticity of rock masses. Our every day experi¬ 
ence tells us that rock is brittle, and the correlative fact of 
its viscosity is not practically accepted on the mere dictum 
of the physicist and the geologist, unless their paths of ap¬ 
proach are to some extent retrod. So results of impact 
which seem to me entirely natural are to some of you ex¬ 
travagant and inconceivable; and if the impact hypothesis 
is to abide with you, it must ingratiate itself by an attractive 
array of accomplish ments. 
The analytic examination of volcanic processes left the 
possibility that the small craters of the moon are maars, the 
results of explosion without eruption of lava; the tidal 
process might perhaps make large craters, but could not 
make small ones. These are the only suggested reactions 
originating in the moon itself which appear competent to 
produce the crater forms actually observed. Taken together, 
they cover all the craters, but they cannot be applied as a 
joint theory without arbitrarily dividing a series the grada¬ 
tion of which is complete as to both size and form. The 
impact theory applies a single process to the entire series 
