600 W. T. JAMES 



ing them must be determined. The indices used in contrasting 

 human constitutions are of little value in studying- dogs. 4 

 Most of them involve leg length, and since in the dog the 

 leg length is a single genetic factor which may be completely 

 dissociated from the body type of the hybrids, it is impractical 

 to use this measurement. Indices based on bodily length, in 

 conjunction with chest width and chest thickness, have not 

 shown significant differences. The cephalic index was first 

 considered as a means for contrasting the dogs, and then 

 the cephalic index in conjunction with bodily length and 

 thickness, but there is no appreciable difference in the cephalic 

 indices among dogs. This is shown in detail in the measure- 

 ments made by Johnson in Section III of this volume. The 

 head shape varies more in the nasal region, the jaws and in 

 length of ears, and width of zygomatic arches, than in the 

 brain case. Absolute measurements of any one bodily factor 

 are not satisfactory. They tell us a great deal about regional 

 development, but little about the total morphological structure. 

 After a consideration of many possible measurements and 

 indices, a relatively simple index, based on the thickness or 

 leanness of the body, was used in this study, even though 

 it was realized that this is not an adequate index for giving 

 all the morphological characteristics of an animal. However, 

 this region of the body is determined by a complex of genetic 

 factors and its index has proved practical in contrasting the 



4 The adequacy of the indices used on the human lias been discussed by Cabot 

 ('38) and Wertheimer and Hesketh ('26). 

 Piget index: stature (cm) — chest circum. (cm) plus wt. (kg). 

 leg length 2 10 



Wigert index: 



Wertheimer-Hesketh index : 

 K. A. P. index: 



chest trvs. dia. chest sag. dia. 



leg length 10 3 



trans, chest dia. gas. chest dia. trunk lit. 



standing lit. 2 

 sitting lit. wt. 



There is some agreement among investigators as to which factors are most 

 significant, but the indices are not as yet definite enough to separate the human 

 satisfactorily. If a number of individuals are compared by different indices an 

 overlapping will be found between types. 



