562 
A Study of the Effects of Diphtheria Isolation 
These are not very considerable, but they are consistent, and indicate, as 
far as they go, that the incidence of diphtheria is not dependent upon such 
measures as the above of unfavourable sanitary conditions. 
If we now turn to the cori elation between the mortality-rate on the population 
and these measures of unfavourable sanitary conditions we find : 
First Period Second Period 
Variates Correlated 1904-1908 1909-1912 
Death-rate from Diphtheria and Infant Death-rate + -081 ± -076 -t- -118 ± -074 
Death-rate from Diphtheria and Overcrowding ... + "061 ± -079 + "004+ "075 
All these are indeed positive, but they are of no significance and if they were 
significant would be so small as to be of no importance. The first indeed might 
have been anticipated to show a higher value, for a certain number of deaths from 
diphtheria must be deaths of infants. We can only conclude that as far as these 
measures of unsanitary conditions are concerned they do not in any way determine 
the diphtheria death-rate. 
We now turn to the isolation-rate and find : 
First Period Second Period 
Variates Correlated 1904-1908 1909-1912 
Isolation-rate and Infant Deathrate ... --414 ±-064 --375 ±-065 
Isolation-rate and Overcrowding ... --236 ±-073 --235+ -071 
These are significant although not very large and we conclude that most 
isolation is practised in those districts which have the lowest infant deathrate and 
the least overcrowding ; the cori-elations are sensible if not very large. In other 
words the towns with better health conditions have adopted more extensively the 
practice of isolating diphtheria cases. 
It seemed further of interest to determine: (i) whether diphtheria and isolation 
were more or less associated with urban conditions, and we took for this purpose 
the number of persons per acre, and (ii) whether the well-to-do character of the 
district, as measured by the number of domestic servants, indoor and outdoor, 
male and female per 100 private families, has any influence on the incidence of 
mortality from, or the isolation of diphtheria. We found : 
First Period Second Period 
Variates Correlated 1904-1908 1909-1912 
Persons per Acre and Attack-rate ... H- "165 ±-075 -f043±-075 
Persons per Acre and Death-rate ... -1- '169 ±"075 -(-•115±-074 
Persons per Acre and Isolation-rate ... -f- '073+ •076 -1- '053 +'075 
Not one of these correlations is of any importance, if indeed any of them can be 
considered significant. It is thus clear that the intensity of urban conditions has 
very little to do with the prevalence of diphtheria, for if anything the suburban 
conditions have the lesser death-rate ; clearly isolation has no sensible relation 
to number of persons per acre. 
