326 
In interpreting the results of these inoculations the following factors 
must be considered, viz: 
(a) The locus of the initial inoculation. 
(b) The virus used in the initial inoculation, 
(c) The locus of the second inoculation. 
(d) The virus used in the second inoculation. 
Bearing these factors in mind we see that Joci chosen for initial inocula- 
tions bear the following relation to the resulting immunity. 
The immunity conferred by skin locus is greater than that by the 
cornea, and the latter is greater than that by the mucous membrane. 
The immunity conferred by vaccine is greater than that conferred by 
variola, as we have already indicated in the preceding section. 
The influence of the locus of the second inoculation can only be esti- 
mated in regard to the cornea and skin when we find that the immunity 
conditioned by the initial inoculation is less efficacious when the cornea 
is chosen than when the skin is the site of the second inoculation. 
The influence of the sort of virus used in the second inoculation upon 
the test seems to indicate that the vaccine virus is more potent than the 
variola virus in that it may produce a lesion in an animal in which 
there is complete protection to inoculation with variola. 
We find, in short, that the skin is relatively a more efficacious locus 
than the cornea, and the latter is more efficacious than the mucous mem- 
brane in producing immunity. The immunizing power of vaccine virus 
is higher than that of variola virus. 
If we interpret the above experiments in the light of the hypothesis 
elaborated in the previous section and keep in mind the physical condi- 
tions at the various loci of inoculation, we feel that the phenomena 
observed are quite consistent. 
A variolation of the cornea after skin variolation succeeds because the 
total amount of immune substance present in the individual is relatively 
small, owing to the character of the virus used in the initial inoculation, 
and because the physical conditions on the cornea do not favor a free 
mixing of the immune-bearing plasma with the inoculated virus. 
In the case of a monkey vaccinated on the skin a vaccination of the 
cornea only rarely succeeds, because the relatively large amount of im- 
mune substance present, even under the adverse physical conditions in 
the cornea, usually is sufficient to produce a germicidal effect upon the 
inoculated virus. We feel that the fact of an occasional animal yielding 
a positive reaction to such a second inoculation only emphasizes our view 
that the phenomenon is a quantitative one. 
In the case of the mucous membrane as a locus of initial inoculation 
the physical conditions are doubtless a large factor in causing the low 
immunizing power. Study of the lesion on the mucous membrane shows 
that almost from the first an open wound is present at the site of inocula- 
tion. This condition would favor the discharge of toxine and products 
