330 
That this incubation period is apparent rather than real, as shown by 
a microscopic study of sections from the inoculation sites, does not help 
us in this connection. By the methods adopted in these experiments 
we are in doubt for 3 days as to the outcome of our inoculation. At 
any time during this period the development of the lesion may be 
checked by the onset of the immunity. We then see that the fact that 
the inoculations on the first 4 days of an experiment result in a diagnos- 
able lesion, while those on the fifth day and on succeeding days do not 
develop to a diagnosable condition, does not fix the date of onset of im- 
munity at the fifth day. Obviously an inoculation done on the fifth day 
might not appear in the records of the experiment as a positive reaction 
if the immunity developed even 2 days later, as the lesion would be 
inhibited before it was recognizable. 
Therefore we must conclude that the onset of the immunity is not 
before the date of the last successful inoculation, and may be as much as 
3 days later. Applying this conclusion to our experiments, where daily 
inoculation was practiced, we see that the development of immunity to 
reinoculation of the skin with vaccine virus may manifest itself at any 
time between the sixth and the eleventh days, and to reinoculation with 
variola virus between the fifth and the eighth days. These figures are 
arrived at by selecting the days of the earliest unsuccessful inoculation 
for the earliest date and the day of the latest unsuccessful imoculation 
plus three for the latest date. These limits, while wide, are as narrow 
as we believe to be warranted by the method of experimentation adopted. 
The appearance of a general exanthem in the animals inoculated daily 
with variola virus, from 3 to 5 days after the last successful insertion of 
virus on the skin, seems at first sight difficult to explain. If we allow 
for a 3-day interval between the invasion of the skin and the appearance 
of the eruptive lesion, we see that this brings the date of invasion, and 
hence the period of intravascular transit of the organisms to within the 
limits set for the onset of the immunity. The number of organisms 
which go to produce a regular, eruptive lesion are undoubtedly infinitely 
fewer than those introduced in an inoculation of the skin with virus. As 
the growth of the lesion depends upon the multiplication of the organ- 
isms, it is probable that the interval between the invasion of the skin and 
the appearance of the exanthem lesion is longer than that between inocu- 
lation of the skin with virus and the diagnosable stage of the primary 
lesion. We must conceive of the organisms which are to produce the 
exanthem as passing from the primary lesion to the skin before the date 
of onset of the immunity. By comparing the dates we can readily under- 
stand how the organisms might make this intravascular journey before 
the immunity developed. Another explanation lies in the possibility 
that phagocytes act as carriers and as protectors of the organism from 
the immune plasma. The development of an exanthem is therefore 
quite consistent with our conclusions with regard to the time of onset of 
