View report page

86
earlier—in the first month of the quarter—and in 1916 slightly earlier again. Similarly in 1919 the
maximum is shown in the middle of the quarter, but it is earlier in 1920. The major " wave " appears
therefore already to have atta'ned its maximum in this epidemic. An exceptionally wet summer
would speed the parting guest, but an exceptionally dry summer might make him linger.
Scarlet fever
and
diphtheria
in relation
to flea
prevalence.
In the annual report for 1917 a summary was given of the evidence collected during the preceding
nine years bearing upon the hypothesis that scarlet. fever is commonly transmitted by the flea (Pulex
irritans). This evidence was considered under ten heads. As regards heads (iv.) to (x.) (with an exception
to be referred to later concerning the relation of flea prevalence to rheumatic fever—a maladv belonging
to Dr. Longstaff's scarlatinal group of diseases—and the bearing of this upon geographical distribution)
there are no fresh facts to adduce, the case stands where it did; a word may, however, be said concerning
the latest experience gained under heads (i.) to (iii.).
(i.) It was noted in 1917 that there was a close correspondence in form between the last nine
seasonal curves of scarlet fever and of flea prevalence, and that the autumnal maximum of the latter
had invariably antedated that of the former, the antedating being by periods of from one to eight or
more weeks. The years 1918, 1919 and 1920 conform with this rule ; the antedating in question last
year was by a period of two weeks. The occurrence of this antedating in each year, over a succession
of 12 years, is a notable phenomenon.
(ii.) and (iii.) relate to the correspondence between the contour of the major wave of scarlet
fever, The corresponding curve depicting annual flea prevalence, and the cycles of dry and wet years.
Points of some interest are to be observed with regard to the most recent developments in this connection.
The crests of the major waves of prevalence were encountered in 1863,1870,1875,1880,1887,1893,1901,
1907 and 1914—the interval between crest and crest being from 5 to 8 years. The wave of scarlet
fever swelled from the trough of 1916-17 during 1919-20. As regards the flea curve a maximum was
attained in 1914, corresponding with the scarlet fever maximum; thereupon followed a decline;
while, however, 1920 showed a distinct rise, the level attained by the flea curve was not equal to that of
1914. The following comments may be made on this : (a) The special attention directed to flea bites
since 1909, i.e., during the last 12 years may be credited, perhaps, with producing some good effect.
(b) But, on the other hand, the rise to the maximum in scarlet fever may not be completely accomplished
until next autumn, and by that time the flea curve may rise to a higher level than that already
reached. It may be noted that the evidence of flea prevalence, forthcoming from common lodging
house inspections, shows a rise continued from last spring through autumn and winter and still maintaining
a comparatively high level even during the first three months of this year (1921). (c) There is
some reason for thinking that the rapid rise in " notified scarlet fever " to its seasonal maximum last
October, may have been somewhat steepened by concurrent development of " influenza " prevalence.
As scarlet fever ran up, a mild epidemic of influenza was steadily gaining ground in London. In some
cases of influenza, especially in children, a rash not very unlike that of scarlet fever may be developed ;
moreover, during November and December, 1920, and January, 1921, " glandular fever " was reported
as occurring in various parts of London (see p. 89) ; and here, again, disturbance of the figures relating
to scarlet fever and influenza was likely to be brought about. The behaviour of the statistics of
scarlet fever and of influenza also will, therefore, be watched with much interest during the next
ensuing months.
It has been urged, by an interested observer of the recent behaviour of scarlet fever, that were
the flea hypothesis true a greater rise in the flea curve should have accompanied tha of scarlet fever
last October. (See Section I of diagram p. 87 ) It needs to be remembered, however, that the flea
hypothesis does not necessitate a precise correspondence between the fluctuations of the flea and
scarlet fever curves. An important influence affecting scarlet fever prevalence, which does not
necessarily involve flea prevalence, is the extent to which a new generation of susceptible young
children, as yet unattacked ty scarlet fever, has grown up since the last considerable prevalence.
Here the low birth rate following upon the date of the last wave crest (in 1914), and the recovery
of the birth rate in 1919 and 1920, must have had some influence. Again, dry and wet cycles do not
manifest their effects upon the fleas and the scarlet fever at precisely the same points of time. In
this connection the long, dry spell of weather in the spring of the present year may have first favoured
flea development and then somewhat later been in part accountable for the delayed fall of the weekly
notifications of scarlet "fever, which was a striking phenomenon during February and March, 1921.
A rise in the flea curve should, however, as a rule, always ante-date a rise in scarlet fever, and this requirement
has hitherto invariably been complied with. It is a matter of some interest to note that
following upon the exceptionally high scarlet fever prevalence of 1893, the notifications of scarlet fever
remained just about the average in the first three quarters of 1894, but no autumnal rise above the
average was observed in the 4th quarter. (See diagram p. 81).
It remains to add a note with regard to the geographical distribution of the human flea,
and of a men.ber of the scarlatina: group of diseases, to wit, rheumatic fever. In heading (ix).
of the ten heads of discussion in the report for 1917, it was stated that Pulex irritans and scarlet fever
both appear to have spread, with increase in facilities of communication from Europe to other parts
of the globe. In a paper contributed last year by Mr. J. Tertius Clarke of Perak to the Society of
Tropical Medicine and Hygiene (Trans. 1919-20, p. 83), entitled " The Etiology of Rheumatic Fever
from a Tropical point of view," it is pointed out that rheumatic fever, and scarlet fever also, are
practically unknown in the tropics. " In the State of Perak during the years 1912-17 inclusive, there
have been 97,914 deaths recorded in a mean population of 548,606, and not one of these deaths has been
ascribed to rheumatic fever." It is further noted that Sir James Cantlie in the year 1899, in the
Journal of Tropical Medicine, was the first to publish the observation that scarlet fever is not encountered