lies entirely below the six necond mark. The inbred curve nev- 

 er rcach.ea a flattened appearajicc but shows ^^reat irretjular- 

 itiea, particularly on the forty-fifth, sixty-firat, ei^jhty- 

 second, ninetieth and ninety-second days. The inbred curve 

 of relearning is more similar to that of the control, but it 

 must be borne in ;iind that the tv/o inbreds and one normal fail- 

 ing to learn are not represented here. From the tv/enty-second 

 day the normal control curve is perfectly flat at 5.3 seconds, 

 all the animals having relearned. Two of the inbreds failing 

 to relearn, their c\irve of relearning remains slightly irreg- 

 ular and above that of the normals. 



In Plate III nay be seen the distribution curves 

 of learning and relearning of both the inbred and control ser- 

 ies for the maze experiment. The time is given in days -- in 

 groups of five for learning, in groups of two for relearning. 

 As may readily be seen, the advantage from the standpoint of 

 time (days required to learn and relearn) lies wholly in fav- 

 or of the normal control series. 



The question arises as to whether the later genera- 

 tions of inbred rats differ from the earlier in the ability 

 to form habits; that is, is deterioration in this ability pro- 

 gressive, even if, as earlier stated, deterioration in relat- 

 ive brain weight after the 4t}'i generation is not. Of the in- 

 bred rats used in the maze, fourteen were from the 6th genera- 

 tion and seven from the 7th generation. In Taole IV is shown 

 a comparative summary consisting of the daily averages of all 

 the 6th and all the 7th generations. From this table, too, 

 are constructed the curves of learning of the 6th and 7th gen- 

 eration inbreds. The table shows that two of the 7th genera- 

 tion failed to learn the maze; all the 6th generation had learn- 

 ed it at the end of eit^hty- three days. The 6th generation re- 



