614 THE AMERICAN NATURALIST. [VoL. XXXIII. 
at the window and the jumping upon the chair; all the rest is a com- 
ment by the observer, and the hypothesis of artifice must maintain 
itself by further proofs than the incident in question affords, against 
the supposition that the acts were independent and both of them 
naive. The parsimonious view is adopted by Dr. Edward Thorndike 
in a monograph,’ which Professor Mills discusses in the present arti- 
cle. Aside from the more special problems of Memory and Imitation 
in the lower animals, concerning which he is at issue with the writer 
of that monograph, Professor Mills takes up two questions of the first 
importance in comparative psychology. The first is that of the meth- 
ods and conditions of experimenting; the second that of the interpre- 
tation of data. 
In systematic investigations of the life of the lower animals it is 
often unavoidable that the subject of experimentation should be 
surrounded by artificial conditions,— range must be limited, dietary 
changed, and daily routine of acts readjusted, but every such 
innovation is a fresh obstacle in the way of the sincere observer, 
and to introduce them wholesale, or to overlook their disturbing 
influence is simply to destroy the whole value of one’s results. In 
many cases the familiarity or strangeness of the environment is the 
controlling element of the experience, and to fail of taking it into 
account is to miss the whole significance of the action. Dr. Thorn- 
dike’s results come very near to being valueless if, as Professor Mills 
says, “ This investigator has practically ignored this in his tests, for 
he placed cats in boxes only 20x15 x 12 inches, and then expected 
them to act naturally. As well enclose a living man in a coffin, 
lower him against his will into the earth, and attempt to deduce 
normal psychology from his conduct.” 
The value of Dr. Weir’s observations lies very greatly in just this 
fact, that his book is the result of a score of years afield, where he 
studied the animals as he found them, living the free life of their 
natural habitat, and not under the inhibitions and disturbances of 
an artificial laboratory environment. 
The second point of Professor Mills’s discussion is of scarcely less 
importance, namely, the interpretation of the data afforded by the 
actions of the animals under observation. Comparative psychology 
labors inescapably under the disadvantage of an indirect method of 
observation. Here, unlike all human psychology, no experience can 
be reported upon by its subject. The investigator must depend 
1 Animal Intelligence. Monograph Supplement to the Psychological Review, 
vol. ii, No. 4, whole number 8. 
