5°4 



THE QUARTERLY REVIEW OF BIOLOGY 



belongs to Thorndike, Kline, and Small, 

 the influence of previous workers, partic- 

 ularly Lubbock, should not be over- 

 looked. As we have already noted Lub- 

 bock originated both problem method and 

 maze method, and used both extensively in 

 his study of insects. In answering the 

 criticism of Mills against his own work, 

 Thorndike (2.1 1) expressly states that 

 "Lubbock used practically the same 

 method with insects" and remarks that it 

 is "odd" that Lubbock's principle of 

 interposing an obstacle between an animal 

 and some incentive such as food, etc. had 

 not been "sooner followed with mam- 

 mals." The influence of Mills is seen 

 clearly also in Small's genetic study of the 

 white rat. As early as 1886, Mills (171, 

 page 3 1) had organized an association for 

 the study of comparative psychology at 

 McGill University and his later observa- 

 tions on both wild and domestic animals, 

 particularly the cat, have served as a 

 model for most later genetic studies. 

 After making due allowance for earlier 

 influences, it must be granted that the 

 task of adapting these experimental 

 methods to the mammals required a high 

 order of originality and ingenuity. The 

 problem situations of Thorndike and 

 Kline were, after all, quite different from 

 those of Lubbock, while the maze of 

 Small was more like the wasp maze of 

 Kline than the simple paper mazes of 

 Lubbock. 



The problem boxes of Thorndike 

 naturally differed from those of Kline, 

 since each had planned puzzle devices with 

 special reference to the particular type of 

 animal to be tested. The general method 

 of the two men differed also in one impor- 

 tant respect. Thorndike confined his 

 animal within the problem boxes, placing 

 the incentive on the outside, or in certain 

 cases utilizing mere escape from the 

 apparatus as the incentive. Kline and 



Small, on the other hand, adopted the 

 more natural method of placing the incen- 

 tive within the problem box, thus leaving 

 the animal relatively unrestrained in 

 working the puzzle device and securing the 

 food inside. The techniques employed in 

 the study of imitation also differed in 

 much the same way. Thorndike confined 

 the imitator in a compartment imme- 

 diately adjacent to the puzzle box requir- 

 ing it to observe the imitatee attacking 

 the problem through a screen or through 

 openings in the side of the box. Kline and 

 Small, and later Hobhouse and Porter, 

 allowed the imitator and imitatee to work 

 at the puzzle device together without 

 restraint. 



The main criticisms urged against 

 Thorndike 's work on the cat and dog by 

 such men as Morgan (178, ziz, review), 

 Hobhouse (143), Kline (153) and Small 

 (109) centered around the alleged unnat- 

 uralness of confining his animals in such 

 small boxes (c. zo X 15 X 12. in.). Mills, 

 with less show of judgment, went further 

 and insisted that valid results could not 

 be secured on the higher animals by 

 laboratory methods. He (171, 173) criti- 

 cised not only the work of Thorndike but 

 also that of Kline, Small, and Hobhouse, 

 all of whom had taken unusual care to 

 make the test situation natural. Al- 

 though neither Small nor Thorndike 

 found evidence of rational imitation in 

 common mammals, the conclusions of 

 the latter were especially singled out for 

 attack, largely because of the contro- 

 versial style in which they were presented. 

 It was quite generally urged also that the 

 negative finding of a single study did not 

 afford sufficient support to Thorndike 's 

 general denial of imitative and rational 

 learning in all mammals below the 

 primates. 



From these beginnings the laboratory 

 study of the higher vertebrates spread 



