J. Blakeman 
147 
noteworthy that Dr Lee's diametral formula* gave a mean error in predictiug skull 
capacity also of about 50 cm.^ It seems to us that no linear formula based upon 
external measurements is likely to give any better result. For all such measure- 
ments are more or less closely correlated, and, as has been shown elsewhere, 
increasing the number of correlated variables does not much narrow the range of 
variation of the predicate •!*. 
TABLE XX. 
Prediction of the Brain-xveiglits of 48 Individuals from 
External Measurements. 
Mean Error, in each instance of 24 cases. 
Formula 
S sex 
? sex 
(i) 
47 '50 grs. 
50-65 grs. 
(ii) 
48-83 grs. 
50-65 grs. 
(iii) 
52-63 grs. 
49-10 grs. 
(iv) 
60-21 grs. 
60-15 grs. 
(c) It may be of interest to try these formulae on one or two actual heads. 
In the Societe d' Anthro'pologie de Paris, Bulletins 1866, II. Serie, T. I. p. 209 
are given the measurements of a skull reputed to be that of Dante. Unfortunately 
all the measurements we require are not given. We find for cranial lengths : 
L = 178 mm., B = 140 mm., U = 525 mm. The bi-auricular transverse arc, doubtless 
measured from centre to centre of auricular orifice, was 310 mm., and the total 
height of the skull, i.e. the basio-bregmatic height, 140 mm. By examining crania 
with a similar transverse arc and basio-bregmatic height, we consider that 118 mm. 
would be a close value of the auricular height of Dante's skull. The height of the 
corresponding skeleton was 155 cms. The mean stature of the Tuscans is 165 cms., 
and of the more northern Piedmontese 162 cms. We shall almost certainly 
exaggerate Dante's stature if we put it at 164 cms. We have next to convert 
skull measurements into head measurements, and in doing this we have rather 
magnified the values in the light of recent investigations on the thickness of the 
flesh. Dante died at 56 years of age. Alongside Dante we can take the measure- 
ments of Jeremy Bentham's head|, and for further illustration the head measure- 
ments of one of ourselves. We have then : 
* Phil. Trans. Vol. 196 A, p. 238 et seq. Biometrika, Vol. in. p. 371. 
t Phil. Trans. Vol. 190 A, p. 466. 
X Biometrika, Vol. in. p. 394. 
19-2 
