312 
Anthropometry of Scottish Insane 
TABLE IV. 
Goodness of Fit. Normal Curve. Individual Asylums. 
Value of P. 
(For o mm. grouping.) 
Aberdeen 
Dumfries 
Dundee 
Edinburgh 
Montrose 
Argyll 
Ayr 
Banff 
Elgin 
Fife 
Glasgow (Gartloch) 
,, (Lenzie) 
Govan 
Haddington 
Inverness ... 
Lanark 
Midlothian 
Perth 
Eoxburgh 
Stirling 
Greenock ... 
Paisley 
Number of very 
crood fits 
L 
B 
H 
S 
Males 
Females 
Males 
Females 
Males 
Females 
Males 
Females 
•916 
■042 
•911 
■164 
■171 
•184 
■831 
•350 
■815 
■797 
•850 
•281 
■869 
•372 
•915 
•782 
•942 
•797 
•893 
■636 
■920 
•447 
■730 
•995 
•986 
■880 
•311 
■221 
■995 
•934 
■219 
•341 
■770 
■973 
•886 
■660 
■925 
•606 
■610 
■860 
•758 
•238 
•273 
uoo 
■042 
•864 
■756 
■834 
•778 
■834 
•822 
•506 
■730 
•180 
■885 
■782 
■998 
•768 
•645 
•860 
•044 
•822 
■977 
•584 
•899 
■954 
•588 
•255 
•807 
•732 
■062 
•413 
•925 
■983 
•516 
•600 
•016 
•296 
■188 
■756 
•629 
■091 
•107 
•234 
•516 
•953 
•624 
■612 
•678 
•719 
•570 
•304 
•144 
•195 
•021 
•183 
•249 
■444 
•819 
•403 
•098 
•382 
•022 
•875 
•720 
•881 
•136 
•801 
■954 
•991 
•612 
•296 
•391 
■943 
•526 
■963 
•085 
•900 
•030 
•333 
•163 
•277 
•948 
■658 
•825 
•169 
•597 
•056 
■690 
•749 
•939 
■277 
•885 
■779 
•782 
•331 
•752 
•834 
•072 
■822 
•214 
■730 
•612 
•842 
•438 
•680 
•995 
■139 
■617 
■837 
■952 
•644 
■885 
•636 
•903 
■9.53 
■701 
■537 
•811 
•423 
•933 
■972 
•964 
•684 
■429 
■901 
•189 
•393 
■639 
■420 
■986 
■576 
•343 
■451 
•875 
•959 
19 
18 
17 
14 
13 
15 
15 
14 
" exceptionals " which are mostly " outliers " in the tables of pairs of distributions 
considered. The second main point to be noted is that the values for males and 
females are approximately equal and do not diverge to the extent shown in the 
values of the corresponding coefficients in most other published results. In the 
" entire insane " series the most highly correlated pair of characters is that of 
Z & jS, and then follow L &l H, B & H, L &i S, H Si S, and B S. The greatest 
divergence between the values of the coefficients for males and females appears in 
the case of the pair of characters L, S, the difference being '3284 — '2573 = '0711. 
There is closer agreement in the "general insane" series, the greatest diflference, 
■0355, between the values for males and females occurring in the case of the pair 
of characters L, H. 
The reader will find in Table V. a summary of the coefficients evaluated while 
the lines of regression of head length on statui'e, $ , and head breadth on head 
length, (/, are shown in Diagrams VIII. and IX. It will be seen from Table VII. 
that, taking head measurements, — to be directly comparable — there is a higher 
