View report page

86
Annual Report of the London County Council, 1911.

N.B.— In some cases persons returned as fish eaters or ice cream eaters have purchased not only at Xs shop
or Y's shop, but also elsewhere. These persons are necessarily included both under the heading " X's shop" (or" Y's
shop") and under "Others," and consequently some of the total numbers of eaters and non-eaters shown in the above
table exceed the total population. The rejection, moreover, of a few indefinite returns causes slight differences in the
total populations shown in the two tables.
An important inference may be drawn from the figures as regards the areas supplied by the
fried fish shop and ice cream shop respectively. The proportion of eaters of fried fish from the suspected
shop, in the families questioned, of course diminishes on receding from the shop (from 68 per
cent, in the "inner radius" to 49 per cent, in the "outer ring"). The proportion of ice cream eaters
from the suspected shop is much smaller to begin with, and diminishes far more rapidly than this
(from 19 per cent, in the "inner radius" to 1 per cent, in the "outer ring,").' It is quite clear, in fact,
that the area from which the suspected ice cream vendor drew his customers is a much smaller one
than the area from which the fried fish vendor drew his customers.
From these figures it clearly follows that an ice cream hypothesis (apart from any examination
into the number of ice cream eaters among the sufferers from typhoid fever) is prima fac'e inadequate
as an explanation of the outbreak ; for nearly three-fourths of the ice cream vendor's customers were
from the " inner radius," and yet more than half the sufferers were in the " outer ring.*'
Further study of the table shows that in the " inner radius " only six fried fish eaters were found
who obtained fried fish from a source other than the suspected shop, all of wrhom, however, also purchased
from X's shop, but in the <£ outer ring " there were numerous fried fish eaters obtaining fish
from other sources than the suspected shop. A difficulty, however, with regard to the fried fish
hypothesis here presents itself. If the control populations be divided into those to the north and those
to the south of the centre, it transpires that nearly half the former give a history of purchasing fried fish
from the suspected shop, and yet the number of cases of typhoid occurring in this population is quite
small, for the brunt of the epidemic fell upon the population south and east of the centre. It was found
later, however, that this comparative freedom from attack of the population north of the centre finds
explanation when consideration is limited to the early evening purchasers (see p. 88). For, when the
middle day and later evening purchasers are excluded, the percentage of customers among the northern
population to total customers of the suspected vendor becomes quite low—in fact is less than 2 per cent.
A further point of considerable importance may now be mentioned. If a fried fish hypothesis
be, for the purpose of argument, accepted, it is possible to predict (from the figures given in the
Table) the number of sufferers from typhoid fever who would probably be found to eat ice-cream.
In the " inner radius " the ice cream eaters supplied by a particular vendor, at ages 5-25, are 42
per cent, of the whole number of persons. It might be expected, therefore, that out of the total number
of sufferers in the inner radius, about half would be Y's customers ; in fact 14 out of 25, i.e., 56 per cent,
were found to purchase ice cream from him. On limiting the enquiry to a sample of the " outer ring "
population lying to the south of the shop, there were at ages 5-25 only 6 per cent, of Y's customers, but
in certain portions of the " outer ring " close to the " inner radius " the percentages rises higher, say,
to from 15 to 25 per cent. Hence, out of the total number of sufferers in the " outer ring," less than a
quarter should be Y"s customers ; in fact, there were only 6 or 7 out of 26, i.e., about 25 per cent. This
analysis demonstrates, therefore, the inadequacy of an ice cream hypothesis on the one hand, and on
the other the adequacy of a fried fish hypothesis to explain the distribution of ice cream eaters among
the sufferers.
Having now disposed of the ice cream hypothesis, it becomes necessary to examine how far a fried
fish hypothesis explains the peculiar topographical distribution, age distribution and distribution of
attacks among families and in single sufferers, observed in the Finsbury outbreak.
Study of the topographical distribution of the cases shows, as has already been mentioned, certain