View report page

88
in the tropics save as an imported infection and never spreads. It is added that " The number of fleas
in the tropics is small compared with the numbers in Europe "and that "The human flea (Pulex
irritans) and the European rat flea (C. fasciatus) have not been found in the Federated Malay States."
Mr. Clarke referred in his communication to the observations on the flea and scarlet fever which have
been made in London, and at the request of the Secretary of the Society of Tropical Medicine and
Hygiene I contributed a note, which appears (p. 89) as an appendix to Mr. Clarke's paper.
Comparison
between the
curves of
seasonal
prevalence
of scarlet
fever in New
York, Paris
and London.
High autumnal prevalence of scarlet fever is the rule in London year after year and it seems
strange, therefore, to find the contours of seasonal incidence shown for New York and Paris strikingly
different from those of this country. In German and Scandinavian towns, however, the English
autumnal periodicity appears to be the rule, at least in those the records of which have so far been
examined.
The amplitude of the yearly wave in New York is far in excess of that for London, and this
suggests the possibility that the density of the populations concerned may have an influence in this
connection. In order to determine this point, the London periodicity has been examined in relation
to " social condition," which is, of course, much the same thing as the density of population. It will
be seen from section 4 of diagram p. 84 that in the denser populations of group 3 the amplitude is much
greater than in group I, It has further been found that there is also a considerable degree of
relationship between the amplitude of the " major waves," or recurrent epidemic periods, and the
density of the population; the relatively severer incidence in epidemic years upon the crowded
populations being followed by years of relatively decreased incidence.
The density of the population has, then, much to do with the amplitude of the "seasonal"
wave; and the conclusion seems to be warranted that the explanation lies in the fact that in the denser
populations the disease runs more quickly through the population and consequently more thoroughly
picks out the susceptible persons, leaving a population of resistant or insusceptible persons, and as a
result a relatively lower incidence of the disease. On the other hand in the less dense population the
infection threads its way more slowly, and a corresponding degree of exhaustion is in this case not
reached.
This explanation does not, however, account for the maximum incidence occurring at different
times of the year in different towns ; but it is possible that here also the density of the population
comes into account, though in a different manner. If the direct operation of climatic conditions be
disregarded, the cause of seasonal periodicity must be one that recurs annually and at about the same
time of the year in any given town. In the diagram (p. 84) showing "seasonal" incidence in London,
Paris and New York, the range of the midsummer holidays is shown. In New York the holidays cover
annually a period of some ten weeks, in Paris eight and in London four and a half weeks. During
these periods, as has been shown earlier, the population is reduced in density ; probably the proportion
of reduction is greater in the more densely populated parts of New York than in those of London. At
this time of the year then the course of the infection through the population is largely interrupted in
New York ; and when the schools reassemble, many of the " threads " of infection have been broken.
In London on the other hand, the interruption is nothing like as great and may not suffice to disestablish
what is perhaps to be regarded as the "natural" periodicity.
The London holiday period of four and a half weeks applies, however, only to children of the
Council's schools ; in the case of private schools the midsummer holidays are considerably longer, and
average perhaps seven or eight weeks. Not only, therefore, would children attending private schools,
as a whole, live in less densely populated areas, but their density would be more largely reduced in
holiday time, and for longer, than in the case of the " Council children." With these facts in mind,
the contrasts shown in section 4 of diagram p. 34 may be in some measure accounted for. In group
I. of metropolitan boroughs in which the proportion of "Council children" of ages between 5 and 13
is least, the incidence of scarlet fever increases from the beginning of the year ; in group 3, the poorest
districts, however, there is a decline to a decided minimum at about Easter time. The proportion
of " Council children " in the least poor districts varies between 60 and 80 per cent., while in the poorest
group the percentage is about 100. Assuming that the " Council children " form 70 per cent. of the
total school children in the first-named districts, and that the seasonal contour relating to them roughly
follows that of the children of the poorest districts (a considerable assumption this) attempt may be made
to arrive at an approximate seasonal contour for the 30 per cent. of private-school children in the
first-named districts. A contour obtained in this manner shows some resemblance to those of
Paris and New York, inasmuch as it now exhibits a period of seasonal maximum about May in addition
to the autumnal maximum. The question as to possible influence of decreased density in association
with longer school holidays in connection with seasonal incidence seems deserving of further study.
There remains to be considered the question as to whether the reduction in the number of
cases of scarlet fever occurring as the result of the midsummer holidays means an actual escape of a
corresponding number of children from attack, or whether the children who escape in the holidays
succumb to infection when the schools reassemble. It is found that there is very little difference in
the total number of cases per thousand children in populations of different social class, and this points
to the probability that those who escape during holidays enjoy only a temporary respite. Susceptibility
to infection appears to be the determining factor, rather than exposure to infection, for the range of
exposure in the densely populated poorest areas must be far greater than in the better-class populations,
yet the total incidence upon the two classes differs but little. The possibility that a large
proportion of cases escape detection and notification in the poorer class populations has, of course,
to be taken into account, but the circumstance to which reference has been made above, that the
distinction between the densest and the less crowded populations lies particularly in the amplitude