68 
Variation and Correlation in Brain- Weight 
The Bohemian females are fairly accordant in variability with the other series 
in both length and breadth, but the males are apparently somewhat more variable. 
This greater variability in the Bohemian male series is due I believe not to less 
homogeneous material but to the manner in which the measurements are made. 
The coefficients of correlation between length and breadth may next be considered. 
Eace 
$ 
? 
Bohemians ... 
•4006 
•1954 
Whitechapel English* 
•240 
•350 
French t 
■089 
- -042 
German! 
■286 
•488 
Enghsh Criminals ^ 
■402 
English Middle Classesf... 
■345 
Ainol 
•432 
•376 
Naqadat 
■344 
•143 
The results here are somewhat chaotic, but clearly the difference in degree of 
correlation between Bohemian males and females is too great to be entirely normal. 
It is probable that the male coefficient is somewhat too high and the female too 
low. The agreement between the Bohemian male coefficient and Macdonell's 
value for measurements on the heads of living English criminals is striking. We 
should expect measurements made as the Bohemian were, to agree better with 
measurements on the living head than with those from the skull. 
The practical result to which we come is that conclusions from the Bohemian 
data as to the correlation between brain-weight and skull length and breadth 
must be only tentative until they can be tested by other less objectionable 
material. 
We may now examine in more detail the results as to the correlation of brain- 
weight with these skull characters. In Table XVII are given the values of r, tj, 
and Sji/. 
TABLE XVII. 
Linearity of Regression. Brain-iueight and Skull Characters. 
Race and Series 
r 
? 
6 
? 
6 
? 
Bohemians, Skull Length 
„ „ Breadth 
•5482 + •0273 
•4655 ± -0306 
•3604+ -0465 
■5041 ± ^0399 
■5660 
■5498 
•4032 
■5368 
•1408 0- 
■2926 a- 
•1808 a- 
•1845 0- 
The four regression diagrams concerned are Diagrams XX to XXIII. 
Macdonell, loc. cit. p. 233. t C. D. Fawcett, loc. cit. 
X Lee, loc. cit. 
§ Macdonell, Biomttrika, Vol. i. pp. 181 and 188, " Measurements on living head and not skull." 
