292 Contributions to the Study of Interracial Correlation 
observed only as averages, and separate races may present many exceptions. The 
cause of the negative correlation between stature and cephalic index is thus only 
one of many factors which can influence the head-form. We must, therefore, 
conclude that the diameters of the head and the cephalic itidex respectively 
depend on many causes. The anthropometric measurements are often artificial 
simplifications, and will, witli the progress of anthropology, be made to approach 
more closely to those regions which are physiologically and embryo logically 
individualized. The oi'ganic influences which exist in the individual, or within the 
limits of a race, can be masked, and correlations between numerical values of 
measurements inverted as we pass to interracial correlation because other factors 
come into play. The one of these factors is sexual selection, which is very powerful 
in the species Homo. Admit for instance that the general growth of an organism 
produces an elongation of the head. Then if the selection is connected with 
stature, and tall individuals survive, the new race thus formed must be long-headed. 
But if round-faced individuals become a type of racial beauty, the sexual selection 
works in the opposite direction, i.e. produces a round-headed tall I'ace, so that in a 
particular case the organic influence can be masked. 
The progress of the evolution of a race is the result of a number of organic 
correlations, and intellectual or instinctive impulses play an important part in it. 
Many correlations which are preserved interracially are organic, i.e. physiological 
correlations. Perhaps we may define the characteristics of a given race as its 
deviations from the most probable values, calculated from the coefficients of organic 
correlation. These deviations are the result of processes which were not the same 
in the evolution of all races, and which have been quite special for each individual 
race. 
