632 
PSYCHOLOGY: R, M. YERKES 
correct box for any group of boxes which may be used as that which 
bears a certain definite spatial or numerical relation to the other mem- 
bers of its group. Definitions which have actually been employed 
(problems presented) are the following: (1) the first box at the left end 
of the group (as faced by the subject) ; (2) the second box from the right 
end of the group; (3) alternately, the box at the left end and the box 
at the right end of the group; (4) the middle box of the group. 
The boxes are presented in varying groups in accordance with a pre- 
arranged plan. The subject is punished by confinement in the box 
selected every time it makes an incorrect choice and is then allowed to 
choose again, and so on until it finally selects that box which is by defi- 
nition the correct one. It is then rewarded with food and permitted to 
pass through the box and return to the starting point, where it awaits 
opportunity to respond to a new group. 
The experimenter keeps a precise record of the subject's choices and 
of various important aspects of behavior. These data include the 
nature of the choices from trial to trial, series to series, day to day; 
the appearance and fate of specific reactive tendencies or methods of 
attempting to solve a problem; and the final outcome, in success or 
failure, of prolonged effort. 
The essential statistical features of the results obtained with certain 
types of subject may be summarized briefly thus: 
(1) Crows quickly solve problem 1 (first mechanism at one end of the 
group), with 50 to 100 trials. 
Problem 2 (second from the end) they fail to solve in 500 trials. No 
consistent improvement appears, although there are four conspicuous 
reactive tendencies: (a) to go to the end box because of previous train- 
ing in problem 1 ; (b) to go to the first box at the left and then to the 
one next in order, which in the particular experiment happened to be the 
correct one; (this is the most nearly adequate tendency exhibited by 
the crow) ; (c) to reenter whichever box happened to be chosen first and 
to choose next the second box from the left (correct); (d) to enter a 
box at or near the right end of a group, and on emerging, to turn to 
the right and enter the box directly in front, and so on until the correct 
box is located. 
In the method of multiple choices, the crow gives no convincing evi- 
dence of ideational behavior. Its general intelligence is clearly indi- 
cated by alertness, keenness of perception, emotional responsiveness, 
and rapid adjustment to various essential features of the experimental 
situation. It appears to be temperamentally ill-suited to the kind of 
task presented by this method of studying reactive tendencies. 
(2) White rats solve problem 1 in from 170 to 350 trials on the basis 
