514 Proceedings of Royal Society of Edinburgh. [sess. 
fallacies from the mixing up of epidemics due to essentially 
different causes can thus be avoided, and the distribution of the 
disease in the city ascertained for any periods which may be 
desired. These maps also give the situation of the hospitals, and 
in addition are divided by parallel lines into transverse squares, and 
also indicate the quarter-mile zones round the different hospitals. 
A careful examination of these maps shows at once that up till 
31st January 1902, though there might he a suspicion that there 
were a few more cases in the neighbourhood of one of the hospitals, 
yet there was no evidence that the general distribution of the 
disease in the city was much affected by any but ordinary epidemic 
influences. The cases for this period were accordingly enumerated 
in the different areas given by the squares, and the form of the 
epidemic calculated for both the north-south and east-west dis- 
tribution of the cases. It was found that the latter was of the 
expected type, namely, type IV., while the former, being very 
asymmetrical, comes under type YI. The theoretical maximum 
of the curves being next calculated, the centre of the epidemic — 
i.e. the locality where, up to this period, the greatest number of 
cases had occurred — was accurately ascertained ; and with this as 
centre, the distribution of the epidemic in zones around it was 
ascertained. 
It was seen at once that the epidemic, from the point of time 
at which it started till the 31st of March, groups itself much more 
naturally around the point which has been found to be the 
theoretical centre of the epidemic than about the hospital, and 
that this centre is at a distance of about one mile and a half in a 
direct line from the area from which infection is supposed to be 
disseminated. It shows at once the fallacy of calculating the 
prevalence of the disease in any one district and comparing with 
the average of the whole town, since, as we see, much the greatest 
number of the total cases occur in a special area, of which the 
zone around the hospital forms an important part. 
The cases round the Priory Road Hospital, probably spread by 
the hospital, thus reduce to a very small proportion of the total 
cases w T hich really have occurred in that zone. That this law of 
the distribution of the epidemic must be taken into account is 
easily seen when the distribution of the cases of the disease in 
