ALCOHOLISM AND BEHAVIOR OF WHITE RATS 281 



alcohol, the subsequent mbreeding has isolated strain differences 

 which mask the effects of the alcohol in the third generation. 

 The great mass of experiments upon artificial selection give 

 strong support to the belief that strain differences are revealed 

 at once when inbreeding is started from a general population; 

 the great variety of characters that have yielded strain differ- 

 ences upon inbreeding leads to the opinion that they may be 

 found even in the ability of white rats to learn a maze. Arlitt 

 admits this point and used random selections of rats in the 

 beginning, "In order to eliminate group or strain differences" 

 (p. 6). The comparison of the second and third generations of 

 her rats affords evidence that such genetic differences between 

 the sets do exist: "The behavior of the third generation closely 

 resembles that of their parents" (p. 41, second sentence). The 

 sets in the third generation that were superior and inferior to 

 normals came from parents that were respectively superior and 

 inferior. If this doubt as to the elimination of strain differ- 

 ences is correct, even accepting all the differences in the 

 averages as though statistically significant, the results cannot 

 be considered significant. The normals are no longer truly 

 controls, since there is involved in the third generation, a second 

 source of difference whose effect cannot be separated from that 

 of the alcohol. Although the final conclusion reached on this 

 subject by Arlitt is in agreement with our own, namely, that 

 alcoholization effects the learning ability of the grandchildren, 

 we do not feel that this agreement really strengthens our evi- 

 dence. It seems entirely possible that a repetition of Arlitt's 

 experiments with the same procedure and methods throughout, 

 and accordingly with the alcohol exerting the same effect, might 

 give a different conclusion, due to the chance selection of a slow 

 strain to be used as normals. 



GENERAL SUMMARY 



This paper deals with the maze-behavior of the second non- 

 alcoholized generation after alcoholic treatment of both grand- 

 parents. The normal controls were the grandchildren of the 

 uon-alcoholized brothers and sisters of the rats that were given 



