48 VICTOR E. SHELFORD 



common in the experience of the species, a better test of its 

 intelligence in terms of what is so named in our own species, 

 than its ability to solve an entirely unnatural problem ? Why 

 should the intelligence of a monkey be indicated any more 

 by its ability to operate locks, than a man's, by his ability to 

 balance himself upon the swaying branches of a tree ? The 

 comparisons may seem crude, indeed they are so, but the matter 

 in point seems well borne out by the evidence suggested above. 

 The error in rating the intelligence of animals solely, either upon 

 the basis of problems which they normally encounter in nature, 

 or those never encountered is not small. Indeed, experimental 

 psychologists have, to a considerable degree, abandoned at- 

 tempts to rate the intelligence of animals. It is still, however, 

 a common practice among zoologists. To understand the be- 

 havior of his animals to the best advantage, the experimental 

 psychologist, should have first-hand naturalistic knowledge of 

 them. The naturalist is even more in need of a knowledge of 

 experimental methods and results. It appears to one not prim- 

 arily interested in either, that the experimental psychologist 

 should be a naturalist and the naturalist an experimental psy- 

 chologist. 



Theoretically, the explanation of the apparent difference 

 between the relations of animals to natural and unnatural 

 problems is very simple. Direct evidence of recent years tends 

 to show that natural selection does not usually, at least, operate 

 upon structural characters in the manner formerly supposed. 

 On the other hand, a rapidly increasingly body of experimental 

 data shows that animals survive or perish under severe conditions 

 in accordance with their irritability and other physiological 

 characters. Irritability is the chief mechanism of survival. 

 Accordingly, where natural selection has been able to operate, 

 responses to stimuli are commonly advantageous; where it has 

 not, they are often disadvantageous and sometimes clearly 

 detrimental. 



Two other points are noteworthy. The sensitiveness of the 

 animals used, to slight differences in rate of evaporation shown, 

 makes errors in laboratory experiments easily possible. Air 

 current in controls were practically abandoned because some of 

 the amphibians detected differences in the different parts of the 

 control cage which could not be detected with the hand, making 



