98 (250) THE VERY HIGH AND VERY LOW INDICES. 
0.31. For the 2nd group we find for 219 total indices 84 high ones, i. e. a 
proportion of 0.384. When in this group eliminating the cases of re- 
versed dominance, in which not any child had a high index, we obtain 
84 : 180, i. e. a proportion of 0.466. For the 3rd group we find 43 high 
indices with total 83 indices, i. e. a proportion of 0.52 and when elim- 
inating also here the cases in which none of the children has a high in- 
dex, we find 43 : 70 i. e. a proportion of 0.61. So the material is more or 
less in agreement with the mendelian expectation. 
Of the low indices for which the index 75 has been taken as a limit, 
we dispose of a small number of cases. When perusing table XI it ap- 
pears that the low indices mostly occur in families where the parents 
have low indices and among brothers and sisters who also have more or 
less low indices. Statistically this is less evident. For the high indices of 
fable X we found as indices for the parents 153.4 — 169.7; for the low 
indices of table XI the parents have the indices 148.4 (141.6) — 160.4. 
It is striking that, having found some dominance of brachycephaly 
over dolichocephaly, in table XI we do not meet with cases where the 
low index is a segregation from high indices. Among the families of ta- 
ble XI there is none in which both parents and most children have 
high indices, whereas a single child has a much lower index. In table V 
p. 22 we have met with cases where of a series of indices the deviating 
one was a rather low one. | 
For the heads with very low indices we have found a great variabi- 
lity among the men (p. 55, tab. b). Examples of a non-hereditary do- 
lichocephalic headform (FRETS 1919, p. 351, THOMA) are probably: fam- 
ily 374 father, family 57 father, family 16 father (or 16a son), family 
90 third son. 
