132 CARNEGIE INSTITUTION OF WASHINGTON. 



These general remarks will serve to show the lines along which further 

 progress may be expected from the reactions of the dogs in the presence of 

 other animals. They also serve to indicate that the differences observed are 

 probably not due to mere irregularities of environment or experience, since 

 they tend to persist and appear in various situations 



Heredity of behavior in mice. — Two races of mice have been used for 

 this investigation, which was carried on by Miss Emilia Vicari. They 

 were a uniform race of Japanese waltzers that had been rendered very 

 homozygous by over 100 generations of inbreeding, and a race of albinos 

 continuously inbred since 1912 from the stock of Dr. H. Bagg, of the 

 Memorial Hospital. This long history assured a purity that made 

 this material of unusual value from a genetic standpoint. Miss 

 Vicari reports: 



Of the two races, the Japs are deaf and have the waltzing habit, with 

 characteristic agitated, nervous movements when not whirling around; 

 they lack vigor and are small in size; they require care in regard to feeding 

 and nesting, and are liable to gastro-intestinal disturbances. The albinos, 

 on the other hand, hear. Not being waltzers, they are normal in their move- 

 ments, and they are larger and more vigorous than the Japs; they are fully 

 tame, with no fear of unusual noises or shadows. The immediate problem 

 was to discover what racial differences in behavior might be demonstrated 

 by their reactions in a simple maze, and then to trace such differences through 

 the generations following the crossing of these races. The maze used (see 

 fig. 4) consists of two successive compartments with right and left exit doors. 

 To reach the food at the end of a trial the mouse has to take the left-hand 

 door in the first compartment and the right-hand door in the second, the 

 other doors being blocked with glass slides. To prevent the formation of 

 scent trails, fresh paper was spread on the floor for each trial and all other 

 precautions required in such experiments, such as uniformity of surround- 

 ings, of age, etc., were taken. The results are based on the number of perfect 

 trials, the number of successive perfect trials, and the average time per trial. 

 The following numbers of mice have been trained: 45 Japs, 75 albinos, 110 

 Fi hybrids, and 40 in the second filial generation. 



The behavior of the parent races as measured by the reactions in this maze 

 shows that the Japs and albinos are very similar when compared on the basis 

 of the number of perfect trials and on the basis of the number of con- 

 secutive perfect trials ; on the basis of time per trial the averages of the albi- 

 nos tend to be lower than those of the Japs. So, in spite of the manifest dif- 

 ferences between the races, it appears that their behavior in the maze is 

 practically identical, although the Japs do not move as rapidly as the albinos. 

 Turning to the Fi hybrids, a surprising result is found: 10 per cent of the 

 mice in this generation made more perfect trials than any parent in either 

 parent race; some individuals excel all those in the parent races in the num- 

 ber of consecutive perfect trials; and the time averages, instead of being in- 

 termediate between those of the parent races, are considerably lower, lower 

 even than the averages for the albinos. This last result is given when each 

 family is considered by itself. The general behavior of these mice distinguished 

 them unmistakably; they were hard to pick up because of their rapid, darting 

 leaps, and scurrying to cover, like wild mice; they were responsive to every 

 little sound, shadow, or motion. Obviously, like the albinos, they could 

 hear, yet they were more sensitive. Physically they are sturdier than either 



