DEPARTMENT OF EXPERIMENTAL EVOLUTION. 129 



the memorj^-maze and in the first half of the multiple-choice training indicate 

 more than chance? If so, does the possession of alcoholic ancestors in general 

 tend to favor the success of rats judged on these criteria? This last question 

 is answered by the comparisons in experiments 23 and 24, in which the tests 

 are better in the maze. In all but two of the 16 criteria, some cases have 

 been found where the tests are better; the two in which they have not been 

 found are in the left-hand problem with the test set-ups in the multiple-choice 

 apparatus when judged by correct choices and in the multiple-choice, memory, 

 when judged by wrong choices. 



"When the averages for individual rats in aU experiments are averaged for 

 each criterion, the sign of the difference is plus in every case, that is, favoring 

 the controls. Many of these differences are not large, and when the test of 

 the probable error is apphed, it appears that only in 6 cases may they be 

 claimed to be statistically significant, that is, probably due to other causes 

 than chance. In only 1 case are the different criteria for the same actual 

 performances significant; this is in the left-hand training proper on the multiple- 

 choice apparatus. There are 2 other cases of differences that are significant 

 on the multiple-choice apparatus, namely, left-hand test set-ups, when judged 

 by the number of correct fu'st choices (when judged by the number of wrong 

 choices the difference is so small that it falls witliin the range of chance 

 variations), and the memory trials when judged by the number of wrong 

 choices. The two significant differences found in the maze criteria occur in 

 the training proper when judged by the number of errors per day (these same 

 reactions of the rats when judged by the time and the number of perfect 

 trials do not show significant differences) and in the memory trials, when 

 judged by the number of perfect trials. 



"At first glance the result seems clear and the conclusion to be drawn fairly 

 obvious — that the alcohohzed 'test' strains are mentally inferior; but, as has 

 been indicated, the correct interpretation can not be finally made directly 

 from these averages. Much detailed study wiU be required before any 

 generahzation can be drawn." 



CYTOLOGICAL STUDIES OF ALCOHOLIZED RATS. 



The study of the cytological condition of the testes of the male 

 alcoholized and the control rats produced in the course of these experi- 

 ments was made by Dr. Ezra Allen, of the Wistar Institute of Anatomy 

 and Biology, and results of the study were published in the Anatomical 

 Record for April 1919. Very briefly, his findings are: 



" There is testicular degeneration in both alcohohzed and normal rats, but 

 this is much greater in the alcoholics." 



In addition to the alcoholized rats, Allen used, for the sake of com- 

 parison, rats that had been raised on a diet deficient in -water-soluble 

 vitamines, but found the same sort of degeneration as in the alcoholized 

 rats. Other investigators have found the same type of degeneration 

 as a result of subjection of the gland to the X-ray. It appears then 

 that similar states of degeneration may arise in the testes of rats 

 through subjection to the X-ray, through deficiency in vitamines and 

 through alcoholization. Allen concludes that the immediate cause 

 affecting growth and cell-division of the germ-cells is identical in all 

 three cases. 



