50 PROCEEDINGS: BOSTON SOCIETY NATURAL HISTORY. 
lected a considerable body of evidence which goes to prove that, 
contrary to the commonly received opinion, men are not more 
variable than women. In another paper, indeed, Lee and Pear¬ 
son (’97) have tried to show that among civilized races men 
are even less variable than women. Neither of these conclusions 
is borne out by the evidence here presented. In the case of the 
Pompeians — the only civilized race for which I have the data 
for both sexes — the men are more variable in 13 dimensions, 
and the women in 5. Furthermore, the mean of 19 coefficients 
for the men is 3.60, and for the same number of dimensions for 
women is 2.68. 
As for the equal variability of the two sexes in general, my 
tables show 60 cases in which the men are more variable and 
36 in which they are less variable than the women. Moreover, 
Table 5, which gives the mean coefficient for each race of Tables 
A, B, C, and D, shows that in each of the 4 cases in which the 
coefficients of both sexes are given, the male is the more variable. 
Lee and Pearson offer evidence (’97) to show that civilized 
races are more variable than savage races. This conclusion is 
contradicted by an equally large body of evidence in the present 
paper. For it may be seen from Table 5 that the civilized 
Pompeians are less variable than the other two savage races, the 
coefficients for which are based on bone measurements; and that 
of the remaining races, the savage Javanese are more variable 
than any of the five civilized races except the Magyars. Thus I 
venture to think that these questions of the relative variability 
of men and women and of civilized and savage races are still 
unsettled. 
