568 A Study of the Effects of Diphtheria Isolation 
table, and there can be little doubt that on the whole sea-port conditions and 
the big new neighbourhoods round London favour, while manufacturing con- 
ditions restrict, the incidence of diphtheria. We have not data, however, available 
upon which we could test water and milk supply, or extent of consumption of 
milk and fish in these towns. The results for Derby and Lincoln are remarkable, 
but they are high for all three periods, and this notwithstanding the rapid 
increase of isolation in those towns. 
At first sight it seemed to us that the towns in the first column were markedly 
those in which there had been a greatly restricted birthrate*, while those in the 
last column were towns of greater fertility. Taking the births per 100 married 
women from 15 to 45 {B) we found : 
r^j, = + 013 ± -075. 
Thus there is no association between incidence and the well-to-do character of a 
town as estimated by a low birthrate. 
Ag'ain havinof regard to the character of the towns in our first cohimn, it 
occurred to us to test the incidence in relation to the employment of males in 
manufacturing processes involving smoke. We took out of the 1911 census the 
percentage {S) of males over 10 years of age, who fell under a rough test of 
smoke-producing occupations, namely ix. 1, x. 1-2, 5-8, xiv. 1, XV. and xvill. 
1-6 of the Registrar-General's classification, and we found : 
r^,, = _-180 ± -073. 
This is possibly significant and would undoubtedly be emphasised had we 
included as a factor the women engaged in textile industries. There seems 
therefore some slight reason to suppose that the conditions favourable to smoke 
production are unfavourable to the spread of diphtheria. 
If the data could be procured, it would be worth while considering water and 
milk supply and the extent of fish consumption in the towns we have dealt with. 
If these were found to be of little influence, the road would certainly be clearer 
for dealing with the chronic diphtheritic human carrier as the chief source of 
the spread of the diphtheria bacillus. 
(10) Conclusions. 
(«) No influence of greater isolation in reducing the attack-rate from diph- 
theria is discoverable. In fact there is a sensible, if not large, positive association 
between the isolation-rate and attack-rate. 
(b) The case mortality is somewhat less where there is more isolation. This 
may very probably be accounted for by more cases coming under specialised medical 
care. 
* We had partially in view here also the possibility that restricted birthrate meant employment of 
women and thus less breast-feeding and greater use of milk, so that cross-currents might be at work. 
