View report page

ratios and those for all ages. This is not surprising because only nine per cent. of male
and seven per cent. of female deaths in 1955 for example, occurred at ages under 45 years.
To sum up so far, the bare fact of long residence in London appears to be likely to
increase the risk of dying particularly from bronchitis; the effect of a smoky atmosphere
is also to increase the risk but while a brief sniff of sulphur dioxide is not particularly
harmful the dual effects of both high pollution and long residence in London are
associated with high mortality. Association is, however, not necessarily a matter of
cause and effect. The reason for the poor numerical association of mortality with the
indices of atmospheric pollution in London may be that the differences within London
are not sufficiently large to demonstrate any difference in the areas shown in Table IV.
It is noticeable, however, that Greenwich and West Woolwich, with the lowest but
one index figure for smoke and the lowest for sulphur dioxide and a moderate figure
for percentage born in London has the lightest mortality experience.
(ii) The short-term effect
In Figure 2 are depicted the daily concentrations of smoke and sulphur dioxide and
the highest and lowest temperatures for four foggy periods since the beginning of this
investigation, together with average levels of the pollution experienced in the two
winters (October to March) of 1954-55 and 1955-56. Temperature has been included
because, as will be demonstrated later, it too plays an important role in short-term
mortality. There is no accepted definition of what constitutes a fog, the severity of the
effects of which will vary according to the local circumstances—a fog in country
districts may be nothing more than mist, a harmless moisture laden atmosphere, but
the same phenomenon when charged with the fumes given off by the domestic fires
and industrial processes of a large city may become a lethal 'smog'. Clearly the fog,
or 'smog', with which we are concerned in this context is one in which atmospheric
pollution is high so that criteria such as a fog service on British Railways or a mention
of fog in the meteorological observations at Kew Observatory are insufficient : clearly
also a 'smog' of sufficient importance to an enquiry of this sort must be reasonably
widespread and long lasting. The criteria adopted, quite empirically, were a mean of the
daily readings at all seven stations in excess of 100 milligrams of smoke
per 100 cubic metres of air and 20 parts of sulphur dioxide per 100 million parts of air,
i.e., about double the customary winter level, on any of two days within a period of
seven consecutive days; in addition the days surrounding these high readings when
there was any mention of fog in the meteorological reports from Kew Observatory
were included. The practice of aggregating the readings of several stations was
deprecated earlier in this survey but in mitigation it can be said these seven stations are
all at or about the same level above ground (first floor level), they are not located for
the express purpose of measuring a particular local source of pollution, and the
measurements are carried out in precisely the same manner at each of them. Not all
the seven stations were functioning throughout the four periods, two were not in
operation during the first period and there were odd days when one or other of the
stations was not in operation, but these minor defects should not invalidate the broad
conclusions. In brief the characteristics of these four periods may be listed as follows ;
a more detailed description is set out in the annex.
Period Intensity of fog Duration Pollution Temperature
Nov. 1954 Moderate Medium Moderate Warm in spite of night frost.
Jan. 1955 Light Long Moderate Very cold.
Jan. 1956 Thick Short Heavy Cold.
Dec. 1956 Light— Long Moderate Cold.
Moderate
In Figures 3a, b, c and d are shown the daily deaths and a seven day moving average
of daily deaths of persons dying in London during periods of about one month of
which the foggy days form the central section. A preliminary analysis was made of
Mortality
in foggy
periods
204