276 
of them seem to belong to the lymphoid-plasma cell series, while some are 
probably of endothelial origin. 
In the older lesions evidence of repair is found. The epithelial cells 
grow in from the epithelium at the periphery of the lesion and upward 
from the ducts of the glands. ‘There is active, new formation of blood 
vessels and of connective tissue beneath the lesion. 
In the portion of the nose covered by columnar epithelium, the process 
is somewhat different from that described above. We find large areas 
denuded of epithelium and a marked reaction of an inflammatory sort in 
the tissue beneath. The lack of coherence between the columnar epithelial 
cells in this situation prevents any marked thickening at the edge of 
the lesion. 
The protoplasmic phases of cytoryctes variole were present in the 
epithelial cells in lesions of all durations and in all situations. Infected 
endothelial cells were demonstrable in one lesion. They occurred in a 
capillary beneath a lesion of six days’ duration on the stratified epithelium 
of the nose. 
DISCUSSION. 
The foregoing experiments show that an inoculation upon the nasal, 
buccal, or oral mucous membrane of the monkey with vaccine virus pro- 
duces a characteristic, self-limiting lesion. Histological study of these 
lesions shows them to be similar to those produced on the skin by vaccina- 
tion. The differences observed in the vaccine lesion on the mucous mem- 
brane from that on the skin are readily explained by the physical condi- 
tions at the site of inoculation. The vaccine lesion on the mucous 
membrane shows no crust or vesicle. The absence of these characters 
is undoubtedly due to the fact that the epithelium at this site of inocula- 
tion does not possess a horny layer, and also to the fact that the surfaces 
are constantly rubbing one against the other and are bathed with fluid. 
We have seen that small collections of fluid between the swollen and . 
degenerated cells of the lesion may take place in the vaccine lesions of 
the mucous membrane. This process is the same as that seen in the 
peripheral portions of a developing vaccine lesion on the skin. However, 
as soon as a considerable amount of fluid collects the tendency is for it 
to escape on the surface, and so no large vesicle forms. The reaction in 
the tissue beneath the lesion is similar in kind to that beneath a vaccine 
lesion of the skin, but the process is less in degree. If we assume that 
the reaction in the corium beneath the skin vaccination is due to sub- 
- stances absorbed from the specific process in the epidermis, it is to be 
expected that the reaction would be less intense beneath a vaccine lesion 
on a mucous membrane. In the latter situation the greater part of the 
products of the process in the epithelium must escape on the surface. 
The mildness of the constitutional reaction and the slight degree of fever 
which these animals present is doubtless due to the latter cause. The 
