specifically infected sewage emptied directly into the river or lake, or by excreta deposited along the banks, which sooner or later are washed into the water supply. By indirect infection we understand those instances in which the excreta, or wash waters, have been deposited in or upon the soil, and by percolation finally reach wells, springs, or other bodies of water.

That the excreta from a single typhoid patient have infected the water supply of a whole community has been shown in connection with the epidemic at Cumberland in 1889; the history of which shows that there were no cases of typhoid fever in that town until the discharges from a young man, who returned from Ohio in December, quite sick, found their way into the public water supply. This case died December 20th. The next case appeared January 10, 1890, and within a short time 485 cases developed in a population of 1,200. Every case but one was traced to polluted river water. The accompanying chart shows that the epidemic resulted in the highest typhoid mortality rate ever reached in this city, viz., 104 deaths in every 100,000 of the population.

The typhoid fever epidemic at Plymouth, Penna., which affected over 1,100 persons, and began April 9, 1885, was traced to a case who contracted the disease elsewhere, returned to his home in January, 1885, and was for many weeks quite ill. The dejecta from this case were thrown along the sloping banks of a mountain stream where they remained innocuous upon the snow and frozen ground until sometime between March 25th and April 1st, when thaws set in. Under the influence of this thaw they were washed into the mountain stream and carried into the reservoir which supplied the town with water. There is every reason to believe that the typhoid fever germs were distributed in the water between March 28th and April 4th or 5th. The first case was reported April 9th. During the week beginning April 12th from 50 to 100 cases appeared daily, and on one day it is said that 200 new cases were reported.

Such instances might be multiplied by the hundreds; we have simply to recall the epidemic at Ithaca, N. Y., during the months of January, February and March, 1903, affecting 1,350 persons in a population of 13,000, with 78 deaths, among them 128 students at Cornell University with 26 deaths. The epidemic at Butler, Penna., in the same year, affecting 1,348 persons with 111 deaths in a population of 18,000; the epidemic at Columbus, Ohio, in 1904, with 1,640 cases and 166 deaths in a population of 140,000; and the more recent epidemic of Scranton, in 1906-07, with 970 cases and 77 deaths, are other shocking examples of water-borne typhoid.

INFLUENCE OF WATER SUPPLIES UPON TYPHOID FEVER RATES. Sanitarians are so convinced that a pure water supply is incompatible with a high typhoid fever rate that they unhesitatingly declare, "Show me a city's statistics of the disease and I will tell you the character of its water supply." The writer, in a paper on "Conservation of life and health by improved water supply," presented a number of diagrams showing the typhoid fever rate in a number of cities, classified according to their water supply, of which plate 7 is here reproduced.

This diagram also conclusively shows that the European cities which do not only prevent pollution of rivers, but also filter their water, enjoy the lowest typhoid fever rates. In the same paper the writer has shown that in seven American cities the reduction in typhoid fever, after the introduction of pure water, amounted to 70.5 per cent, and that during the past 25 years the death rate from typhoid fever in 14 countries and cities, tabulated by him, has been reduced 54.3 per cent. The Bulletin for the month of April, 1908, of the New York State Department of Health shows that the death rate from typhoid fever in 10 cities of that State has been reduced 53.4 per cent by an improved water supply.

CAUSES OF TYPHOID FEVER.

The writer has heretofore held that about 50 per cent of the cases are water-borne (well, river, commercial and ice), and has taught for years that wells in this city, on account of soil pollution, should not be tolerated, and that the only way to render a suspicious supply safe for drinking purposes was to boil the water. The danger from Potomac water, since filtration, has been practically eliminated, but ice should not be mixed with it as it may be as impure as the water from which it is derived. Moreover, iced drinks and foods are also objectionable on account of the injury to the digestive functions, especially in the summer months, unless sipped very slowly, and, upon general principles, ice cream should be made with sterilized cream.

In the writer's report on typhoid fever for 1895, he urged as a preventive measure the boiling of milk, and the systematic inspection of dairies. This, together with the fact that he has tabulated 195 of socalled milk-borne typhoid epidemics, clearly indicates that in his judgment milk is to a certain extent responsible for the dissemination of the disease, probably in about 15 per cent of the cases. In 148 of the 195 epidemics analyzed by him there is evidence of the disease having prevailed at the farm or dairy. In 67 instances it is probable that the infection reached the milk by soakage of the germs into the well water with which the utensils were washed, and in 16 instances the intentional

MEAN DEATH RATES FROM TYPHOID FEVER, 1902 TO 1906, IN 66 AMERICAN CITIES AND 7 FOREIGN CITIES. GROUPED,
AFTER FUERTES, ACCORDING TO THE QUALITY OF THEIR DRINKING WATER.

RATES FOR FOREIGN CITIES ARE TAKEN FROM JAMES H. FUERTES.

THE