Karl Pearson 
255 
xjax = + "5567. 
Here* x,Ict, = + -Qbld, ^./(t., = + -2934, x,la, = + -bbod, 
xja, = + -9811, x,la, = + -4844. 
The correlation ratio for this table is = 'IBS, but a little examination shows 
that there is no definite association of poor health in child and alcoholism in 
father. Of the healthy children and of the delicate children the fathers have equal 
intensity of alcoholism ; the phthisical and epileptic children come in an over- 
whelming proportion from the sober parents. The parents of those who die young 
are somewhat more alcoholic than those who survive. The small ditference may 
well be due to more accidents and not to worse health in the children of alcoholic 
parents. 
(F)bis. Health of Sons and Alcoholism of Fathers. Miss Elderton's table is 
as follows : 
Health of Sons (Manchester Data). 
Healthy 
Fair 
Delicate 
Phthisical and 
Epileptic 
Died 
Young 
Totals 
Alcoholic 
122 
9 
24 
8 
42 
205 
Non-alcoholic... 
328 
37 
71 
37 
133 
606 
Totals 
450 
46 
95 
45 
175 
811 
Heref cc/o-j; = + '6658, and for the arrays: 
*i/(Ti = + -6095, x,la-o = + -8573, xja, = -6062, 
xJa, = + -d2Sd, xfa, = + -7063. 
The correlation ratio is here r) = -0H9, but the correlation coefficient obtained from 
a fourfold table of healthy against the remainder is — '06, or the children of alco- 
holic fathers are the more healthy. It is quite easy to see from the above results 
how little evidence there is for alcohol in fathers affecting the health of sons. As 
in the case of girls, but more sensibly so, the healthy cljildren have fathers some- 
what more alcoholic than the average ; the delicate children have fathers not less 
sober than the average; the phthisical and epileptic children have fathers markedly 
more sober than the average population, and the early deaths are slightly more 
frequent among the children of sober parents. It is difficult to see in either of 
these tables evidence for alcohol as a chief source of bad health or poor intelligence 
in the offspring. 
(5) As a last illustration I take certain statistics of the severity of attack in 
cases of smallpox and the strength of immunity provided by vaccination. I owe 
the following data to Dr John Brownlee of the City of Glasgow Hospital. 
* Reduced to a common standard : Si/o-,^^ + -5470, x.,I(Fj.= + ■290S, ^3/0-^.= + -.SolO, ^4/0-^= + -9725, 
56/iT^= + -4801. 
t Reduced to a common unit: xi/o-,^— + •G071, l2/<^j;= + ■8539, 53/0-^.= + -6035, 24/0-5.= + -9202, 
^5^= +-7035. 
33—2 
