that the general contour of the two curves is broadly traceable to common causes, there remains
a possibility that some effect may be exerted upon diarrhoea prevalence by flies. In this connection
the following considerations need to be borne in mind.
It is generally assumed—to quote words used by Dr. Longstaff 30 years ago—that there is a
specific cause of diarrhoea, such as "some animal or vegetable matter in a state of change, or even
a living germ." This cause may itself be subject to meteorological influences, and may, therefore,
operate more intensely at one time of year than another. On the assumption that the fly acts as
carrier, the prevalence of diarrhoea must be determined not merely by the distribution of flies, but also,
and, indeed, more particularly, by the seasonal intensity of operation of the specific cause. This fact
is apt to be lost sight of by those who hold that, because there is a fairly precise correspondence in
time of flies and diarrhoea, there is therefore a causal relationship between them.
If, on the other hand, as some maintain, almost any organism associated with putrefaction may
communicate to food materials the ability to cause diarrhoea, it should be noted that the quantity and
quality of germs floating in the air are largely influenced by season, and it may be found that the
seasonal curve for "totalgerms" presents correspondences with the curves both of flie3 and diarrhoea.
If this should be demonstrated to be the case, the role of the fly in spreading diarrhoea may prove to be, at
most, only a subsidiary one.
(6) Comparison of the seasonal curves of " intestinal diseases " and of flies.—It has been claimed
that the fly has been largely concerned in some instances in spreading typhoid fever; indeed, in
America the insect has been designated the " typhoid fly." A difficulty which presents itself in this
connection has been felt both here and in America. Thus, in Dr. Daniel D. Jackson's report on the
"Dissemination of Intestinal Diseases through the agency of the Common House Fly," a chart is given
showing weekly deaths from diarrhoea and from typhoid fever in relation to temperature. The typhoid
fever deaths have, however, been set back two months, and the curve thus brought into correspondence
with that of diarrhœal deaths. It is explained that this has been done to cause the curve " to correspond
to time of contracting the disease." It is difficult to justify so long a period as two months being fixed
upon. Inasmuch as both the typhoid fever and diarrhoea curves relate to deaths, an antedating of some
five weeks only would appear to be indicated. The same discrepancy is manifest in this country, as
Dr. Peters has shown (Trans. Epidem. Sect. R.S.M., 1908-9, pp. 5 and 43). Dr. Peters, who it must
be noted is dealing with diarrhoea deaths and typhoid fever notifications, says, " the beginning of the
typhoid is seen to average about four weeks later than the diarrhoea, or about three weeks if a reference
be made in each case to the date of infection." Thus, in America and in England, the time of maximum
distribution of infective material is some three weeks later in typhoid fever than in diarrhoea.
This fact is not reconcilable with a thesis that the seasonal curves of diarrhoea and typhoid fever
are determined solely by extent of prevalence of a common carrier, viz., the fly. On behalf of the fly
hypothesis it might, however, here again be urged that the specific causes of both diseases themselves
respond to seasonal influences—the diarrhœal harvest maturing first, then three or four weeks later that
of typhoid fever. Although there is not simultaneity, spread of infection may, in both cases, be facilitated
by the fly, just as flowers of different species, maturing] some earlier some later in summer,
are fertilised by insects.
(c) The decline in jxrevalence of diarrhoea in relation to diminution in number of flies.—In a
previous report it was pointed out that the London diarrhœal curve of 1907 showed marked decline
at a time when the number of flies was still excessive, and reference was made to the observation of
Dr. Niven, Medical Officer of Health of Manchester, to the effect that " the descent in the number of
deaths precedes the diminution in the number of flies." Further experience confirms these observations.
Thus the London deaths, in the early weeks of September, 1909, showed a distinct tendency
to decline, and the disease must presumably, allowing for incubation, etc., have shown a similar tendency
at least a week or two earlier, i.e., at the end of August. Yet throughout the first half of September
flies were much more numerous than in the first week in August when diarrhoea was rapidly increasing.
Dr. Niven has suggested that the phenomenon may be due to exhaustion of susceptible material.
But it is observed not only in years of high mortality, when such exhaustion might have influence,
but also in years of very low mortality, when it could not operate to any appreciable extent. The
fall of diarrhoea, in the first fortnight of September, commonly observed in an ordinary season, in
London to-day, occurred also, allowance being made for change of style, in Sydenham's time. That
author writes : " This disease, however much it may be epidemic, very rarely occurs beyond the limits
of the month of August, the month in which it begins. This makes me admire the beautiful and
subtle mechanism of Nature in the determination of the origin and decline of epidemics, since even,
although the same causes, viz., the abundance of fruit" (and he might have added, the prevalence of
flies), " are common to the months of September and August, the effect is different." It has, however,
been suggested that the actual gross number of flies is of minor importance, and that activity and
habits, in relation to visits to infected material and to food, must more especially be considered. Thus,
Dr. Nash says (Journal of Hygiene, September, 1909, p. 158): " The numerical abundance of flies
visiting food and more particularly milk, directly after or within a few hours of settling on objectionable
matter, whether out of doors or within a dwelling place," determines the amount of diarrhoea. He
adds : " Although flies may appear to be quite as numerous in houses after the diarrhœal curve begins
to descend, when the temperature of the air falls perceptibly they are less active in their habits,
and do not peregrinate to the same extent in consequence of the cold, but make for the warmer
corners of the room and pay less attention to feeding." This description would, doubtless, hold
good in late October, when autumn is well established, but in 1909 our own observations, made
directly on this point, show that during late September and early October flies were particularly