SUMMARY. (253) 61 
prepotence of the indices of parents through personal or through ancest- 
ral influence. 
Grouping the indices of parents in classes, it appears, that the vari- 
ability of the indices of children of the class of parents with average 
indices is a little higher than of the class of parents with high, resp. 
with low indices. When this result holds, it is a support for the Mendel- 
ian conception of heredity of headform. 
_ Speaking of the different tables we observed, that of families Sun 
some brachycephalic children often one or two children have a very 
high index and a small head. 
The statistical study of the very high and very low indices (tab. X 
and tab. XI, fig. 10 and 11) could not bring to light this difference; 
of the brachycepalic large and the brachycephalic small head there 
is not two-toppedness in the curves for the length and the breadth 
of heads with a high index. It is found, that the heads with high 
indices are upon an average somewhat smaller than those with low 
indices. 
Among the heads with high indices there are more small ones fren 
among the whole material. 
In our material the length of the heads with high indices has de- 
creased in the same proportion as the breadth has increased, certainly 
for the men, for the women the decrease of the length is somewhat 
larger than the increase of the breadth. 
When calculating the variability as a coefficient of variation (p. 55) 
we find somewhat other results than we find for the standarddeviation. 
The coefficient of variation of the headbreadth is in all cases somewhat 
larger than that of the headlength. 
For low indices among men the heads are relatively somewhat more 
long than narrow. The standarddeviation of the headlength is here 
larger than that of the breadth, both for men and women ; head length 
and head breadth of dolichocephalic men have a very large variability. 
