Miscellanea 
253 
P from the tables (Biometrika, Vol. I. p. 155) is about '000,03. In other words the odds are 
more than 33,000 to 1 against the occurrence of two such divergent samples of hair colour 
if they were random samples from the same population. I think we can conclude that they are 
really differentiated samples, or that scarlet fever and measles do not attack indifferently all 
individuals whatever their hair pigmentation ; or, that scarlet fever and measles are differential 
in their selection. (I suppose, of course, that no personal equation actually existed between the 
pigmentation recorders for measles and scarlet fever.) 
Illustration 11. In the Scarlet Fever Cases, Dr Macdonald has added together the Belvidere 
and Ruchill contingents, the measles being based solely on Ruchill returns. I next asked 
whether this might be the source of the differentiation observed. The following table gives the 
work, when we take Ruchill scarlet fever and measles cases only. Thus comparing the two 
diseases at one hospital, we find 
X 2 = 661 x 1295 x -000,0301=25-77, 
or the improbability is precisely the same. The observed result is therefore not due to the 
combination of the Belvidere and Ruchill observations. 
Hair Colour. 
Black 
Dark 
Medium 
Fair 
Red 
Totals 
Scarlet Fever 
Measles 
(i) 
(ii) 
8 
0 
214 
85 
749 
367 
251 
184 
73 
25 
f 
f 
1295 
661 
(i) + (") - 
(i) /1295 
(ii) /661 
(iv)-(v) ... 
(iv)-(v) 2 ... 
(vii)/(iii) ... 
(iii) 
(iv) 
(v) 
(vi) 
(vii) 
(viii) 
8 
•0062 
•oooo 
+ -0062 
•000,041 
•000,0051 
299 
•1652 
•1286 
+ -0366 
•001,340 
•000,0045 
1116 
•5784 
•5552 
+ -0232 
•000,538 
■000,0005 
435 
•1938 
•2784 
- -0846 
■007,157 
■000,0165 
98 
•0564 
■0378 
+ -0186 
•000,346 
•000,0035 
f+f 
fIN 
fW 
flN-f'UT 
{flN-flNJ 
(flN-flN'f 
f+f 
1956 
1-0000 
1 -oooo 
o-oooo 
■000,0301 
Illustration 111. One further point occurred to me as worthy of investigation in 
Dr Macdonald's data. To what extent is it desirable to club Boys and Girls together in 
an inquiry of this kind? 
For Scarlet Fever in Table III. p. 23, we have : 
Hair Colour. 
Black 
Dark 
Medium 
Fair 
Red 
Totals 
Girls 
Boys 
(i) 
(ii) 
3 
5 
137 
77 
389 
360 
138 
113 
38 
35 
/ 
/' 
705 
590 
(i) + (ii) - 
(i)/705 
(h)/590 
(iv)-(v) ... 
(iv)-(v) 2 ... 
(vii)-r(iii) ... 
(iii) 
(iv) 
(v) 
(vi; 
(vii) 
(viii) 
8 
•0043 
■0085 
- -0042 
■000,0176 
•000,0022 
214 
•1943 
•1305 
+ -0638 
•004,0704 
•000,0200 
749 
•5518 
■6102 
- -0584 
•003,4106 
•000,0046 
251 
•1957 
T915 
+ -0042 
•000,0176 
•000,0001 
73 
■0539 
■0593 
- -0054 
•000,0292 
•000,0004 
/+/' 
fix 
fIN' 
fjN-f'IN' 
(f/ir-f/jry 
{fIN-f'IN'f 
~ f+f 
1295 
1-0000 
1-0000 
00000 
•000,0273 
