Donaldson, Brain and Spinal Cord of Rat. 



159 



weight of the body, and the type has been already discussed in a 

 previous paper (Donaldson, '08, p. 350). 



In this connection, however, there are some points to be corrected 

 and further discussed. The consideration of these points is taken 

 up in a paper by Dr. Hatai which apj)ears at this time. 



TABLE 2. 



Mean l)()(ly leiij^tli :u-ci)rdiiig to body weight in Mus iiorvogicus var. albus. 

 The body length as given in the last coUnnu has been calculated by the 

 fomnila (4), y — 143 log (x + 15)— 134. 



The co-efficient of correlation between the body weight and body 

 length, the records being grou})od as stated al)Ove, is found to be .90. 



It is possible, therefore, to infer the body weight from the 

 stature, and vice versa, provided the body weight is normal. 



At the same time it is evident that body weight is much more 

 open to fluctuations than is the body length, and therefore the body 

 leuo'th is the better standard. 



