ence between uody \veit,ht in {_rara8 -md body len^t>i in lai] 1. ime- 

 tres is slij/ht. But under conditions of ovei^feedint^ or of sic'K- 

 ness the body weight varies j^reatly while the bociy lent^th re- 

 mains constant. "For this reason Pr. JJonaldson has accepted 

 body lent^th as the sole criterion. In these tables, however, 

 both nre produced. Eoth body Icn^jth and body v;eit:ht of the 

 inbred rats used in the nane are, on the avera^-e, a trifle 

 greater than is the case with the noriaals. The avera^ie actual 

 brain weight of the inbreds is less than that of the normals. 

 The relative brain weight (in reference to body length) of 

 the inbreds is 4,43;1 less than that of the normals. The rela- 

 tive brain weight (in reference to body weight) of the inbreds 

 is 1.99% less than that of the normals. The percentage of wa- 

 ter in brain and cord decreases v/ith age, but in the inbreds 

 of the maze series, although killed on the aTerage fourteen 

 days later than the normals, the percentage of water was great- 

 er. 



The tables of comparative surru.iaries of mar>e results 

 support the hypothesis that the deterioration of brain weight 

 in a strain of rats is accompanied by deterioration in the 

 ability to form habits. 



In Plate II is shown the curve of learning (below) 

 and of relearning (above) of the inbred rats compared with 

 those of the normal control. The inbred curve is shown by the 

 solid line, that of the control by the broken line. The ordi- 

 nates give the average time in seconds, and the abscissae the 

 number of the day in which such time was made. The time requir- 

 ed by both inbreds and normals for the first four days w.-.s 

 80 long that it is represented here by fig\jrea rather tzian 

 by the curve. For the first few days the descent in both cas- 

 es is very rapid, and from the twentieth day the normal curve 



