F. M. Turner 487 
TABLE VIII. 
Vaccinated cases only, 1901 — 2. Nmnher of Vaccination Scars and Severity. 
Recoveries 
Deaths 
Totals 
Less than 4 scars ... 
.3449 
506 
3955 
4 scars or more 
2761 
169 
2930 
Totals 
6210 
675 
6885 
;•= --2631. 
TABLE IX. 
Vaccinated cases only, 1901 — 2. Foveation of Scars and Severity. 
Recoveries 
Deaths 
Totals 
Scars less than half foveated ... 
Scars more than half foveated... 
5203 
980 
602 
58 
5805 
1038 
Totals 
6183 
660 
6843 
r= --1664. 
These figures agree well with those given by Macdonell for the Glasgow 
epidemic of 1900-1, less closely for the epidemic of 1892-5 in the same town. 
The slight differences in the total numbers observed in the three tables are due to 
the measurement of area or the counting of the number of scars having presented 
special difficulty in certain cases. 
An important practical point to settle, if possible, is whether these three 
correlations are independent or two of them secondary to the third. It seems 
probable d priori that the protection afforded by vaccination is proportional to the 
area of the vaccine vesicles, which may be regarded as laboratories for the produc- 
tion of an antitoxin. If this were so the correlation between number of scars and 
severity would be secondary to that between number and area of scars. 
TABLE X. 
Vaccinated cases only, 1901 — 2. Number and Area of Scars. 
Area < | inch 
Area > h inch 
Totals 
Less than 4 scars ... 
1526 
2398 
3924 
4 scars or more 
165 
2761 
2926 
Totals 
1691 
5L59 
6850 
r=-|--6648. 
Denoting by r^s> ^'ns> ^^/v> the total correlations between area of scars, number 
of scars and severity of disease taken in pairs, and p^^s> Pns> the partial correlations 
