814 
Anthro2)ometry of Scottish Tusane 
degree of correlation in the pair of characters L, B among the "general insane" 
than among the 3000 English cnminals or the 1000 middle class English, which 
show the lowest degree of the three classes. For the pair LS the values are very 
similar for " Entire Insane " males and English Criminals. The values of the 
TABLE V. 
Goeficients of Correlation. 
"Entire Insane" Population 
4436 Males 
3951 Females 
Head Length and Breadth 
Head Length and Height 
Head Breadth and Height 
Head Length and Stature * 
Head Breadth and Stature* 
Head Height and Statui-c* 
•5026 + -0076 
•4027 + -0085 
•3761 ± ^0088 
•3284+ •0091 
•2002 + ^0098 
•2340 ±-0096 
•5235 + ^0078 
•3566 ± ^0094 
•3474+ ^0095 
•2573+ •OlOl 
•2211 + ^0103 
•2357 ±^0102 
"General Insane" Population 
4381 Males 
3925 Females 
Head Length and Head Breadth ... 
Head Length and Head Height ... 
Head Breadth and Head Heiglit . . . 
•4848 + ^0079 
•3755 + ^0089 
•3529 + ^0090 
•4672+^0084 
•3420+ ^0095 
•3325 + •ooge 
corresponding coefficients for some series of skull measurements are given in 
the table. They show marked differences from the series of head measurements. 
In Table VI. are given the values of the correlation ratio, r], and also their 
differences from the corresponding comparable values of (i.e. those found without 
TABLE VI. 
Test of Linearity of Regression. 
" General Insane " Population. 
Males 
r 
'/-'■ 
•67449 - ^ ^ 
Head Length and Breadtli ... 
Head Length and Heiglit ... 
Head Breadth and Height ... 
Head Lengtli and Statiu-e ... 
•4612 
•3547 
•3390 
•3226 
•4573 
•3546 
•3299 
•3133 
•0039 
•0001 
•0091 
•0093 
2-94 
•41 
3^83 
3 '78 
Females 
Head Length and Breadth ... 
Head Length and Heiglit ... 
Head Breadth and Height ... 
Head Length and Stature ... 
•4372 
•3241 
•3109 
•2489 
•4358 
•3216 
•3084 
•2488 
•0014 
•0025 
•0025 
•0001 
1-62 
r94 
\m 
•17 
* 4401 Males and 3915 Females. 
+ Since (ri-r)lr is small this formula gives a good arithmetical approximation to the value of 
(t; - r)jE. See Biometrika, Vol. iv. pp. 348, 341). Blakeman on "Linearity of Regression." 
