468 Journal of Comparative Neurology and Psychology. 



served at the St. Marylebone Infirmary was used for a special 

 purpose. For this last material I am indebted to Professor H. 

 H. Donaldson, who very kindly placed at my disposal his 

 manuscript copy of Marshall's original tables which were nev- 

 er published in their complete form. 



An account of the methods in which the weighings were 

 made in the different series, together with a discussion of the 

 objections to which they are open, is given in the complete 

 paper. 



The usual biometric methods were followed in the deter- 

 mination of the various constants. Each series was treated 

 separately. For reasons which need not be entered into here 

 it appeared advisable to deal separately with two different age 

 groupings of the material. One of these groups included all 

 individuals falling in age between 20 (in the Marchand series 

 15) and 80. This forms what I have called the "Total" group. 

 The other group which I have designated the "Young"' group 

 includes those individuals falling in age between 20 (in the 

 Marchand series 15) and 50. It will be understood that 

 throughout the paper "Total" and "Young" refer to these age 

 groups. 



The other characters studied besides brain-weight were 

 stature, age, body-weight, skull length, skull breadth and the 

 weight of the cerebrum. The units in which the constants are 

 given are : brain-weight in grams, stature in centimetres, body- 

 weight in kilograms, skull length and skull breadth in millime- 

 tres and weight of cerebrum in grams. 



2. Brain-zveigJit Types. A study of the means and their 

 probable errors shows that there are definite racial types in this 

 character. The differences between different races in respect 

 to weight of brain are only in part to be accounted for by 

 differences in other characters of the body. Two of the races 

 studied (Swedish and Hessian) are sensibly alike in mean brain- 

 weight. These are the two races out of the four studied which 

 on other grounds are thought to be most closely related ethnic- 

 ally (cf. on this point Ripley ('99) and Dkniker (:oo) ). That 

 differences in mean brain-weight in different races are not to be 



