154 Biometric Constants of English Brain-weights 
(9) Oh the Human Prime in the Case of Bniin-weight. 
Pearl cites* Marchaiid as stating that the human brain-weight reaches its 
maximum between 15 and 20 years, and remains practically constant until 50, 
when senile degeneration begins. Gladstone cites Boyd and Vierordt to the effect 
that the maximum brain-weights are to be found in youth. Gladstone's own data 
certainly show this maximum in almost extreme youth, but the degeneration begins 
immediately the prime is reached. It is true the data are very scanty, but such as 
they are, they are suggestive. We have : 
Males 8—19, Brain-weight 1407 grs. 
„ 20—25, „ 1394 grs. 
„ 25—46, „ 1380 grs. 
„ over 46, „ 1320 grs. 
Even among the 8 to 19 group the group 12 to 15 has the largest average 
brain-weight, i.e. 1443 grs. The numbers are very small, and of course the cases 
of chronic illness in the older groups are more numerous. But Table I, p. 125 
shows that even among the acute cases the fall in brain-weight owing to age alone 
is very substantial. Accordingly the human prime in brain-weight seems to fall 
before 20 years, and after 20 years there is, on the whole, a continuous fall in brain- 
weight. This result must be steadily kept in view when any attempt is made to 
correlate brain-weight and intelligence. It can hardly be argued that the human 
intellectual prime corresponds to the human brain-weight prime. And while 
brain-weight steadily changes, first rising and then falling with age, intelligence 
will be shown in a forthcoming memoir by one of our number to be almost stationary 
during the rise of brain-weight. Further, it is clear that the correlations and 
regressions will be much modified if we include in our results the rising period of 
brain-weight with the falling period. For this reason we have limited the material 
of this paper to cases in which the individual was 20 years and upwards. In this 
way our regressions will be much more nearly linear. A glance, for example, at 
Powys's diagram for change of stature with agef would, we think, convince the 
reader that without the use of skew correlation curves it is not possible to mingle 
observations on a physical character for ages before and after the "prime." We 
think that the source of the irregularity in the Hessian correlations for weight, 
stature, and age may be due to including individuals between 15 and 20 years 
of age^. 
The determination of the " prime " in every anthropometric measurement is a 
matter of very great importance and interest. Our own experience is that up to 
the prime we get a curved line of growth not widely divergent from the logarithmic 
in character. After the prime the degeneration is very gradual, and for practical 
purposes almost linear. It would be simple, therefore, to allow for this by age 
* Biometrika, Vol. iv. p. 20. t Biometrika, Vol. i. p. 47. 
J Stature has a later prime than brain-weight, and this might well account for the irregular Hessian 
results in Pearl's Table IX. See p. 44 of this number of the Journal. 
