324 
Smallpox and Vaccination 
compared, not only when the risk of infection is comparatively slight, as in 
the general population, but when it is much more considerable, as in infected 
houses. In the annexed table are given in parallel columns a series of coefficients 
expressing the relative susceptibilities under each of these conditions for all age 
periods. It will be seen that the protection which vaccination affords against 
attack is much less strong, considered in relation to the total numbers of such 
persons in the town generally, than to the numbers in houses actually infected. 
In the former it ceases to be considerable after twenty-five years of age, and is 
practically negligible after the age of fifty. In infected houses, however, this 
decline is not nearly so marked, but after the age of fifty it also becomes 
practically negligible. This is an example of a property which at times seems 
to ensure a greater value to action on a rnass than on a small section of that 
mass which differs from the mass in being apparently more vulnerable. It 
is not a case of the poison acting equally on the two classes as may be taken 
for granted is likely to be the case in invaded houses, but that the relative 
sparseness of unvacciuated persons affords them a protection beyond what random 
selection would give. It is conceivable that some such factor plays an adjuvant 
part in natural selection. There are no other statistics which enable comparison 
to be made between those in infected houses and those of the general population, 
but the relative susceptibility between the two classes of vaccinated and unvacci- 
uated in infected houses has been calculated for some other epidemics, Leicester, etc., 
which show local differences, but exhibit on the whole the same kind of relation 
from age period to age period which has been noted in the case of Sheffield. 
All the figures agree in making a rapid fall in the protective power of vaccination 
after the age of fifteen years. It is, however, to be noticed that the protection 
against a severe attack of smallpox or death afforded by vaccination is of a much 
more permanent nature than that against an attack of the disease. This is to be 
expected when the clinical resemblance between vaccination and the second stage 
of the course of smallpox both as regards local appearances and the course of the 
pyrexia is recalled, and when it is noted that the clinical course of smallpox 
suggests the presence of two very different classes of toxins in the two different 
stages of the disease, so that the immunity against the first stage is much more 
readily lost than that against the second. That this is the explanation will be 
apparent when the close relationship between the protection against severe attack, 
that is an attack with a severe second stage of the fever, and a fatal attack is 
noted. There are no very ample data to determine this comparison, but in the 
annexed table will be seen in parallel columns the correlation betwixt recoveries 
and deaths, and mild cases and severe for the Gloucester epidemic of 1896, and 
it will be seen that the correspondence, considering the small numbers involved, is 
sufficiently close. The figures which Macdonell has already given are quoted 
for comparison, when it will be readily seen that the deaths give a very fair 
measure indeed of the severity of the disease, closer than might be expected 
when it is considered how large a factor the personal equation plays in the tabula- 
tion of degrees of severity. This is important, as the great majority of statistics 
