E. TSCHEPOURKOWSKY 
163 
We see firstly, tliat the coefficients lie between '726 and "929, and in the mean they 
are equal to '826 (Table I.). Secondly, that in sexual resemblance there does not 
exist a large difference between characters which are apparently indiffei'ent and 
those which can be presumed to play an important 7-dle in the struggle for life. 
This is shown also by the means of indices and relative lengths, which are nearly 
the same for both sexes. Two explanations of this fact are I think possible : 
1. The racial proportions of the human frame have no special value in the 
struggle with environment, or they have the same value for both sexes. 2. The 
characters acquired in this struggle under the influence of natural selection by 
one sex are inherited by another. 
If we admit that the racial characters of the human frame are evolved under 
the influence of an environment which is different for each sex, then why are they 
not characteristic for one sex only, i.e. why in the species Homo is the male not 
different from the female, as in many other species of animals and plants ? It 
shows I think that in the formation of the different races of man many characters 
are probably the result of homogamic and endogamic instincts. From this point 
of view a race is very clearly defined by Karl Pearson as a population which by 
geographical, religious, or other causes has been isolated during a time sufficient 
for thorough intercrossing and for the work of selection. If that be so, the degree 
of resemblance between male and female of the same race is the product of this 
selection. Within the limits of one race the degree of resemblance between the 
closest relatives of the opposite sex is nearly '4. If a part of the members of this 
race with a special deviation of any one character become isolated and form a new 
race, then the resemblance between their male and female elements increases to '8. 
The difference '4 is perhaps the measure of the work of selection, and the sexual 
correlation coefficient is the criterion of the formation of a new race. 
From the Table I. it follows, that judged by coefficients of intei'racial variation 
woman is in the mean less variable than man ; but the difference is very small. 
In three characters out of seven she is more variable than man. But in at least 
five of the seven cases the difference is not sensible, having regard to the probable 
error of the difference. 
21—2 
