BREWSTER : VARIATION AND SEXUAL SELECTION IN MAN. 47 
score for this test, recorded as before, is given in column 10, and is 
4-2 for both sexes. Column 11 gives the total score for the three 
tests which have been applied to the data of Table A. Altogether, 
in eleven cases out of fourteen, conspicuousness is associated with 
variability. 
In Table B are given the coefficients of various skull and face 
dimensions of Venezuelans, and in Table 2 I have brought together 
averages corresponding to those of the New England Indians given 
in Table 1. If the same comparisons are made among these aver¬ 
ages as were made among those of Table 1, in every case the more 
conspicuous dimensions are seen to be the more variable. 
The first ten columns of Table 3 repeat, for the data of Table C, 
the tests already applied to the data of Tables A and B. Column 3 
shows seven correct cases out of a possible nine; column 6, twenty- 
one out of twenty-seven; column 10 shows nineteen out of twenty- 
seven ; a total of 47 out of 63. All the tests thus far applied give a 
total of 86 to 19, or something more than 80 per cent. 
Suppose now, in Table C, I consider the face measurements alone, 
and divide them into two groups, one of which shall contain 
the coefficients of all the more conspicuous dimensions, and the 
other the coefficients of the less conspicuous. Concerning the con¬ 
spicuousness of some of these dimensions I am unable to form an 
opinion. But I think I am tolerably safe in claiming that the dis¬ 
tance from the ear to the nose root, the distance between the 
temples (upper face breadth) and the distance between the angles 
of the jaw (lower face breadth) are less conspicuous than the under 
jaw length, cheek breadth, and mouth breadth. I may state at this 
point that I made the division first, and computed the mean after¬ 
ward. Column 11 gives the mean coefficient of the three less con¬ 
spicuous dimensions, and column 12 the mean for the more conspic¬ 
uous. In each of the nine cases the relative magnitude of the 
coefficient is what would be predicted. 
I may go even farther, and compare four conspicuous with four 
inconspicuous dimensions, by including with the conspicuous dimen¬ 
sions already taken, the distance from ear to chin, and with the 
inconspicuous dimensions, the distance between the inner angles of 
the eye (nose root breadth). These two dimensions are not so 
clearly assignable to their respective classes as the six dimensions 
first compared. Nevertheless, the distance from ear to chin meas¬ 
ures the protrusion of the lower jaw, one of the most conspicuous 
