F. M. Turner 
TABLE XXIX. All Gases, 1901—2. 
495 
Recoveries 
Deaths 
Totals 
Under oD 
5244 
911 
6155 
Over oO 
2786 
718 
3504 
Totals 
8030 
1629 
9659 
r=+-1360. * 
Whence we deduce 
?>5 = --4246, r^y= + -67l, r^,. = + -13C0, p,.^s = - '7022. 
The discrepancy between the above three values of pyg is largely due to the 
want of linear correlation between age and severity of disease as measured by 
death or recovery. Another possible cause has occurred to me, which may have 
contributed to this result ; and as this cause would also account for the small 
value shewn above for the diminution of vaccination protection with time, I am 
inclined to lay stress upon it. 
Let us take any individual out of a population exposed to smallpox. Three 
things may happen to him, either he may escape the disease altogether, he may 
contract smallpox in a mild form and eventually recover, or he may contract it in 
a severe form and die. Into which of the three classes any individual falls, 
depends on his power of resistance to the disease or his immunity, and on the 
amount of virulence of the infection to which he is exposed. In some diseases, as 
diphtheria and tetanus, we are acquainted with immunity produced by a chemical 
substance, and which may therefore be quantitatively measured. In diphtheria 
and tetanus the above three classes correspond to definite quantitative amounts 
of antitoxin. In smallpox we have not this knowledge, but the analogy makes it 
justifiable to regard immunity as a quantity, though not as yet measurable. The 
immunity may be natural, or acquired by vaccination, or the result of the two 
causes combined. 
If in such a population we desire to examine the correlation between severity 
of disease, which is the observed fact rendering the immunity visible to us*, with 
any other factor as vaccination, age, &c., we may make a sixfold table, thus : 
TABLE C. 
State as regards Disease. 
t-i . 
bo o 
to .2 
!Z2 
No Disease 
Disease and 
Recovery 
Disease and 
Death 
Totals 
Un vaccinated ... 
Vaccinated 
a 
d 
h 
e 
f 
/ 
s 
t 
Totals 
I 
in 
N 
* The influence of exposure to different degrees of virulence of infection is probably also an 
important factor, but is ignored, because at present beyond our observation. In dealing with a large 
population there will probably be no marked differences in this respect between the vaccinated and the 
unvaccinated classes. 
