1 88 1.] Geology as a Science and an Art. 453 
careful allowance to be made for any variations in tempera- 
ture, but, so far as I know, it does not strictly hold good in 
the case of any single known gas. At great pressures all 
gases show deviations from this law ; but this is only be- 
cause they are not under all conditions perfect gases . For a 
perfect gas, could such be found, the law would hold good 
for all pressures. Coming now to the first class of science 
(in the order in which I have enumerated them), we may 
take Zoology, or the natural history of animals, as an 
example. Naturalists study the animal world as they find 
it ; and Mr. Darwin’s law of Natural Selection has for its 
object to formulate how varieties and species are being formed 
under the conditions of our Earth, not how they would be 
evolved if those conditions were different. Botany, Soci- 
ology, Astronomy (at least under one of its aspects), and 
our special subject Geology, also come under this' class of 
science. 
Now all these three kinds of science are mutually helpful 
to each other ; and now-a-days sciences of the first class, 
such as Geology, cannot get on at all without the aid of such 
sciences as Physics and Chemistry, which belong to the 
second class. Nor can the sciences of the second class get 
on at all without the aid of the sciences of the third class, 
Mathematics and Logic. 
The next question is, Are these three kinds of science 
equally exact ? To this I answer, No, they are not. The 
third, with Mathematics for its type, is by far the most exact. 
Pure mathematics is, indeed, the type of exactness and defi- 
niteness. Then comes Physics, the most exact science of 
the second class. Physics we may say is quite exact, wher- 
ever we may be sure of isolating its problems and getting rid 
of, or allowing for, all disturbing elements. Then it can 
apply mathematical processes with perfect confidence. But 
when we come to problems belonging to the first class of 
science, to the study of phenomena as wholes, then we can- 
not and must not expect the same exactness. This is not 
because the phenomena are not subject to law, for all phe- 
nomena whatsoever are subject to law, but because the con- 
ditions are so complicated, and there are such a number of 
influences at work, that the human mind is inadequate 
completely to unravel them. Even in Astronomy, where the 
conditions are by far more easily unravelled than in any other 
concrete science, the theory of the moon is still an unsolved 
problem. 
I dwell on this distinction between these three kinds of 
science, partly because I believe the distinction to be in itself 
