John Brownleb 
327 
be at a maximum between twenty and thirty years, and then to gradually 
disappear towards old age. The figures here, however, are not large enough to 
base any definite conclusions. 
It now falls to be considered at what rate immunity develops aftei' vaccina- 
tion, that is to say, how long must vaccination be performed prior to the onset 
of smallpox to prevent the attack? The incubation period of smallpox is from 
twelve to fourteen days, and it is found to be practically an absolute rule that no 
person develops smallpox more than fourteen days after a successful vaccination 
has been performed. If vaccination be performed during incubation period of 
smallpox the attack is seen often to be much modified. Here again the statistics 
refer to small numbers. In the first place, taking all ages together and estimating 
the correlation between the attacks and deaths in un vaccinated persons and in 
TABLE XII. 
Correlation between Recoveries and Deaths in Unvaccinated Persons, and 
those Vaccinated at various Dates prior to onset of illness. 
Time before 
Gloucester 
London 
London 
onset 
(all ages) 
(all ages) | 
'ages above 1 year) 
Days 
1896 
1902 
1902 
0— 3 
■13 
•06 
•40 
3— 6 
•16 
•12 
•40 
6— 9 
•65 
•47 
•72 
9—U 
•14 
•12 
•45 
U~ 
1^00 
TOO 
1-00 
those vaccinated after infection, we find that even among those vaccinated within 
three days of the onset of the illness there is a distinct positive correlation in 
favour of the process. Among those vaccinated from three to six days before the 
onset of attack the correlation is slightly higher, from six to nine days it is very 
considerable, from nine to fourteen days, however, it is again very much less, 
apparently due to the fact, as will be shown later, that only a few persons who 
are really very susceptible develop the disease after nine days. After fourteen 
days the protection against death is absolute. If children under one year be 
eliminated from the numbers considered the protection is seen to be considerably 
higher, but still to exhibit a maximum at the period of six to nine days. 
Though the subtraction of those cases under one year makes such a difference 
in the correlation coefficients, yet when those who have been vaccinated more 
than six days before the onset of the illness are compared for the varying age 
TABLE XIIL 
Correlation between Recoveries and Deaths of those Unvaccinated, and those 
Vaccinated 6 — 14- days jmor to onset of illness. London, 1902. 
Age Periods 
0— 1 
•61 
1— 5 
•6.3 
5—10 
•73 
10—15 
•38 
15—20 
•49 
20—30 
•61 
•60 
