350 Journal of Comparative Neurology and Psychology. 



manifest imitative behavior, the complexity of the problems can 

 be increased and thus by successive steps the range of imitative 

 ability can be determined. 



Viewing the matter in this way, I deemed it important to give 

 the monkeys an extended pi'eliminary study. I was unable for 

 some time to set problems which seemed well suited to the purpose, 

 and my best ideas seemed to come accidentally as I was observing 

 the- animals. From a large number of possible problems, selection 

 and combination was made so that, in the end, I had a group of 

 devices presenting situations adapted to the monkey's ways of doing 

 things. The value of this preliminary work, T am sure, is evident 

 in the results of the experiments. 



2. Laboratories and Apparatus, (a) Laboratories. The experi- 

 ments made in Cambridge were conducted in a research room of 

 the Harvard Laboratory. The living cages were located against 

 the wall of the large airy room. Between these living cages and 

 the experiment cages, a curtain Avas drawn while the experiments 

 were in progress. Light fell upon the experiment cages from two 

 large windows so that all parts of the a]>paratus were well illumin- 

 ated. The room was on a third floor and on the side away from 

 the street. It was, therefore, exceptionally free from the noises 

 and jars of traffic. 



At the Park, a laboratory was arranged in a room at one end of 

 the Primates' House. The room was 15 feet long and 12 feet 

 wide, with good light from two sides and the roof. Along the two 

 sides of the room opposite the windows, were the living cages, where 

 the animals were kept, two and three in a cage. Between these 

 cages and the windows a floor space, 7 by 10 feet, gave sufficient 

 room for the experiment cage described below. Its wire sides were 

 toward the windows so that it might be well lighted. The experi- 

 ment cage was separated from the living cages by curtains which 

 could be drawn back Avhcn the experiments were over. 



(6) Experiment cages. In presenting problems to inonkeys one 

 meets two difficulties at once. If the animals are left free in a 

 room they wander about, examine everything in the room and 

 give only intermittent attention to the problem, thus wasting time. 



