498 Proceedings of Royal Society of Edmburgh. [sess. 
shown. One example of this — an outburst of yellow fever * in 
Demerara — is given (diagram XVI., Table A, 32). 
This forms the general survey of an epidemic. The applied 
curves seem to give a very fair approximation to the facts ; but 
when, as in an epidemic, a disease propagates itself amid a variety 
of evanescent and manifestly independent influences, it is evident 
that, whatever the law of its spread, only an approximation can be 
Diagram XIV. 
expected in the result. In the first place, there is the weather, 
exercising an influence which at present there is no means of 
measuring. Secondly, there is the distribution of the population 
in the infected districts. In a city, for instance, there are large 
areas where there is no population, or only a very sparse one, and 
the disease might initially appear in one of these. Thirdly, there 
is the nature of the housing or grouping of the people, as is seen in 
* Harty, Yellow Fever in Guiana , London, 1820. 
