486 Correlation hetimen Vaccination and Smallpox 
From these two tables and a further table of Islington cases, brought up to 
the end of the epidemic, he concludes that the correlation diminished as the 
epidemic progressed. 
I have calculated the correlation separately for the two years 1901 and 1902. 
TABLE V. 
London, all Gases admitted to M. A. B. Hospitals during the year 1901. 
Recoveries 
Deaths 
Totals 
Un vaccinated 
253 
119 
372 
Vaccinated or doubtful 
1198 
173 
1371 
Totals 
1451 
292 
1743 
r= --.3851. 
TABLE VL 
London, all Gases admitted to M. A. B. Hospitals during the year 1902. 
Recoveries 
Deaths 
Totals 
Unvaccinated 
1272 
634 
1906 
Vaccinated or doubtful 
5307 
703 
6010 
Totals 
6579 
1337 
7916 
,■=-■4330. 
I therefore conclude that the correlation did not diminish with the progress of 
the epidemic. Table V. is strictly comparable with Table B in period, and differs 
from it chiefly in taking account of the ultimate result of all those cases remaining 
in hospital at the end of the year. Both my tables include the doubtful cases 
which are omitted in Macdonell's. The efiect is to lower the values of r, but not 
as can be seen by comparing Tables I. and IL to an extent sufficient to account 
for the difference. 
I have also calculated the coefficients of correlation between severity of disease 
and number, area and foveation of scars from the figures for the vaccinated cases 
only ; severity of disease being estimated as in the previous tables by the fact of 
death or recovery. 
TABLE VII. 
Vaccinated cases only, 1901 — 2. Area of Vaccination Scars and Severity. 
Recoveries 
Deaths 
Totals 
Area < ^ inch . . . 
1402 
293 
1695 
Area>§ inch ... 
4784 
379 
5163 
Totals 
6186 
672 
6858 
?•= --2997. 
