84 
when the promixity between susceptible persons becomes 
sufficiently close for the infection to pass freely from one 
to the other. 
Exanthematous diseases rarely attack the same indivi- 
dual twice in his lifetime. 
When, therefore, an epidemic has, by either a fatal or non- 
fatal attack, cleared away nearly all the susceptible persons 
in a population, mostly infants and children up to a certain 
age, then it must necessarily wait a certain number of 
years before the requisite nearness of susceptible individuals 
has been again secured. 
There must in the interval be a gradual re-stocking of 
the nation with material fit for the epidemic to feed upon, 
and it can only spread when the requisite proximity is 
attained, when the meshes of the network of communication 
are sufficiently close for it to include all susceptible persons 
in one grand haul. 
It will be found I think that this hypothesis is supported 
by the peculiarities of the different epidemic curves, and 
especially by the contrast between the courses followed by 
these diseases in England and Sweden respectively. 
1. We may note the extreme height of the epidemic 
waves in the sparsely peopled country of Sweden, as com- 
pared with the comparatively small excess of mortality in 
epidemic years in the dense populations of England and 
Wales. 
Thus the death-rates from measles and scarlet fever during 
an outbreak in Sweden are usually at least ten-fold those in 
ordinary years, and in the case of measles they sometimes 
rise to twenty-fold the number. 
In England and Wales, on the other hand, scarlet fever 
never sinks below 7,000 deaths, nor rises above 24,000 — a 
ratio of less than one fourth — and measles is still more 
certainly and equally present amongst us, the epidemic 
never marking double the number of deaths of ordinary 
