DEPARTMENT OF GENETICS. Ill 



heights the dog begins to whine before jumping, then to cry, then to chatter 

 or bark, and finally to combine all these sounds in a violently excited 

 behavior. As the limit of his j umping is reached the sounds reach a maximum ; 

 as the limit is passed the noises diminish in intensity and the dog may lie 

 down at the back of the box. There are many variations from this general 

 behavior and many individual idiosyncrasies in the character of the sounds 

 made. 



"Some dogs would jump from the elevator when raised to 12 feet; others 

 would refuse to jump at 4 feet. There was no indication of correlation 

 between such differences and the size of the dogs. The tests gave higher 

 limits when repeated after two months, so, in spite of the brief duration of the 

 test, an element of learning was obviously involved. One litter showed 

 great uniformity in jumping from considerable height; another litter showed 

 great irregularity, some making high jumps while others refused to jump 

 much above the initial 2 feet. These seemed to be favorable indications for 

 the success of the method. It had the great advantage of being brief, it gave 

 direct numerical data, and it required no preliminary training. But when 

 the chows were tried, difficulties were encountered; they preferred to sit in 

 the box rather than use the energy to jump. One of them would jump when- 

 ever some extraneous sound or motion attracted her to the ground ; the others 

 sat contentedly in the elevator when it was raised above the first positions; 

 two would jump when called or at the sight of a person. Although it is 

 possible to show that the dachshunds jumped from greater heights than the 

 chows under the same given conditions, such a comparison can have little 

 significance, for it was obvious that the dachshunds were anxious to jump, 

 and in most cases would jump till they bounced high on landing in the straw, 

 while the chows would lie quietly till the proper incentive appeared and would 

 then leap at once with the greatest ease. 



"Whereas the method at first appeared to offer many advantages and to 

 give a simple measure of temperamental differences, it soon became clear 

 that behind its simple data there were many causes. The very same data 

 could mean very different things. When a dog refused to jump from 4 feet 

 it may have meant (1) that the dog was scared by the height, or (2) that he 

 had no interest in the ground, and remained in the elevator because he was 

 contented to do so. When a dog jumped from 10 feet it may have meant 

 (1) that he had a great desire to run about on the ground, (2) that he was 

 scared by the unusual elevation, or by the elevator itself, or (3) that he had a 

 well-knit body, or a light body, that had not been hurt by the jolt in landing 

 upon the straw in the previous jump. One could be led to believe that the 

 height data measured courage, by observing many of the dachshunds and 

 hearing their noises as the height increased; yet, for the chow, height certainly 

 does not measure courage, and quantitatively it probably measures nothing 

 at all. More likely in this case the data indicate that chows were not stimu- 

 lated by a situation that did stimulate dachshunds, but such a difference 

 between the breeds applies only roughly, for reversed conditions were found 

 in both breeds. So by the same units we may be measuring one character 

 in some cases while in others we may be measuring a different character or 

 measuring nothing; and comparisons mean nothing if the data do not measure 

 the same thing. Although considerable data have been gathered from these 

 elevator tests, they are now believed to have little value for a comparative 

 study of native temperament in different families or different breeds. The 

 method is too simple for an analysis of a vastly complicated situation." 



