Death Rate per 100,000

Broadly speaking, there seems to be no particular difference between the typhoid fever rate shown in the Washington and the Richmond cases. There are some specific cases where very high figures are shown in Richmond, notably in 1880 and 1881, when there seems to have been an epidemic. In 1884 there was a very heavy epidemic indeed. In the year 1900 the Richmond curve shows likewise an epidemic. Apart from these special dates, the Richmond curve practically follows the Washington curve, and the typhoid fever rate is almost the same throughout. It is particularly noticeable in the later years since the Washington filters were added. Since this time a marked improvement in

FIG. 3.

1880 81 82 83 84 85 86 87 88 891690 91 92 93 94 95 96 97 98 99 1900 01 02 03 04 05 06 '07 '08 '09 1910 11 12 13 14

TYPHOID FEVER DEATH RATE

(Including Typho-Malaria)

1880-1914

typhoid fever has been found at Washington, and a similar improvement can be found in the city of Richmond. The improvement in the city of Richmond was largely caused by sanitary measures, particularly that of instituting a rigid milk inspection. The date of the starting of this rigid milk inspection is noted on the diagram.

It is not an unfair inference to say that a large part of the reduction in typhoid fever in Washington was owing, not to this improved water filtration, but to other sanitary measures which were introduced at about the same time.

There is one important factor, however, which differentiates the effect of an unfiltered water supply from the effects of a filtered water supply. An unfiltered water supply which may be subjected to occasional gross pollution may produce at intervals a very serious epidemic. Such an epidemic may easily run the typhoid fever rate to several times the normal for a period of a year. Such an epidemic can hardly occur with a properly-filtered water supply. In this lies the particular value of filtration. Typhoid fever as a whole may or may not be reduced in any particular locality by the introduction of filters. In some cases there has been a marked reduction of typhoid attributable only to the filtration and to nothing else. In other cases, as at Washington, such reduction as has been noted is probably not to be attributed chiefly to the filters, but to other sanitary measures.

It is not to be asserted as a general rule that filters must eliminate the typhoid fever and other similar sicknesses in any town. It can, however, be asserted without qualification that such filters will practically eliminate any chance of a typhoid epidemic, and thus make the water supply safe. A surface water supply without filtration or sterilization or prolonged storage cannot ordinarily and uniformly be considered safe. Even under the best of conditions and under the most favorable inspection the water of flowing streams may at times be subjected to pollution which may cause typhoid epidemic.

Dr. A. C. Houston, of London, in his recently-issued book "Studies in Water Supply," considers the question of the relation of water and disease, and from this book we quote several sections, as follows:

"There can be no question that in a number of American cities improvement of water supply has antedated the decrease of mortality, but in our present state of incomplete knowledge to conclude from this circumstance a direct causative connection between these two is perhaps hardly warrantable in all cases.

"The fact that there has likewise occurred a decline in the incidence of tuberculosis, pneumonia, and respiratory diseases generally, none of which can reasonably be regarded as at all likely to be water-borne, throws some doubt on the hypothesis that an improved

water supply necessarily tends to produce any marked and continuously operating decrease of mortality from 'general and particular diseases.

66

The author is, of course, not suggesting, in the face of conclusive evidence to the contrary, that serious typhoid epidemics have not resulted from the consumption of specifically polluted water, or that any relaxa

FIG. 4.

JAN. FEB. MAR. APR. MAY JUNE JULY AUG. SEP. OCT. NOV. DEC.

DIAGRAM SHOWING

IDEAL FOREIGN CONDITIONS

AND RECENT IMPROVEMENTS WHICH HAVE BEEN

MADE IN AMERICAN CITIES.

120

110

100

90

80

70

60

MUNICH +1888, 1897

ANNUAL DEATH RATE PER 100 000

YEARLY TYPHOID FEVER DEATH RATES.

tion in the processes of water purification is permissible; he does venture to express the opinion that the evidence brought forward in favor of the direct causative relationship between the quality of water supplies and the degree of incidence of endemic typhoid fever is not always wholly convincing.

"It is impossible, of course, to accept any theory which associates diminution of the typhoid fever rate as attributable to improved water supply, but seems to ignore increased incidence of the disease as of water significance."

Fig. 4, modelled after Jackson's data in his Cleveland report of 1913, shows some yearly typhoid fever death-rates FIG. 4. (Concl.)

JAN. FEB. MAR. APR. MAY JUNE JULY AUG. SEP. OCT. NOV. DEC.

under various conditions for a number of cities. The first section, in the upper left-hand corner, gives a diagram showing certain foreign cities (Munich, Dresden, and Hamburg) with actual annual typhoid fever death-rates per 100,000, running in all cases below ten and averaging about five. The conditions represented by these figures are practically ideal. It is well

worthy of note in these lines showing death-rate for various months during the year that incidence of typhoid fever is practically constant during the whole year.

The second section of the diagram, in the lower left-hand corner, shows improvements effected in two cities by general sanitation. The heavy line represents the city of New York. The upper line shows the New York typhoid for the years 18871896; the lower solid line, New York conditions for the year 1911. There has, during this time, been very little, if any, change in the water supply; in any case, not a very serious change. The improvements in typhoid fever death-rate must be looked for in other directions.

The two Boston curves are shown by the dash line: the upper one, Boston from 1888 to 1897; the lower one, Boston in 1911. The improvement here is very much more marked, and is also to be explained largely by improvements in general sanitary conditions, although partly by the Wachusetts water supply, introduced in 1898.

The next part of the diagram, the upper right-hand corner, shows combined improvements effected through change in water supply and general sanitation. It is not possible very exactly to divide the effect of the water supply improvements from the effect of a general sanitation improvement. In some cases, notably that of Cincinnati and Philadelphia, there is not much doubt that the improvement in water supply explains the greater part of the reduction of typhoid fever. In the cases of Cleveland and Chicago it is quite possible that improvements in water supply played a somewhat smaller part in the reduction of typhoid fever death-rate than believed by some. However, these supplies were grossly polluted at intervals of short duration in earlier years.

The last set of curves, in the lower right-hand corner, shows the typhoid fever rate for four cities during the years noted on the curve. The particular feature to be noted is the extreme range of typhoid fever during the various parts of the yearlow during the months of January and February, and very high during the summer months. The inference to be drawn is that the water supply, which does not change noticeably during the year, is to blame for only a relatively small part of the typhoid, and that the greater part is to be explained by other causes, such