2o6 THE POPULAR SCIENCE MONTHLY. 



iug water could hardly receive a more striking demonstration. 

 Yet the possibility has been realized in the experience of Melun, 

 a garrison town of about twelve thousand inhabitants, situated on 

 the Seine, twenty-eight miles above Paris. Here, in 1889, there 

 were one hundred and twenty-two cases of typhoid fever among 

 the soldiers. The Chamberland filters (Pasteur system) were 

 then introduced, and the cases of the following years numbered, 

 respectively, fifteen, six, two, seven, and seven again for 1894. 

 Suddenly, during the severe weather of February of this year, 

 twenty-eight dragoons, one after the other, came down with the 

 fever. The infantry battalion, living in the same barracks, had 

 not a single case. The secret was soon out. The filters had been 

 allowed to freeze and the soldiers were ordered to drink only the 

 weak infusion of tea furnished them, in which, of course, the 

 water was boiled. The dragoons had simply not obeyed, but had 

 helped themselves to the Seine water from the hydrants. 



At Lorient, as in other districts of the coast of Brittany, ty- 

 phoid fever has long been endemic and still remains so among 

 the civil population. In the garrison, until 1889, there was an 

 average number of one hundred and seventy cases yearly. In 

 1890 filters were set up, with the result of a decrease in the cases 

 to fifty-eight for that year, while the three following years num- 

 bered only two, one, and one, respectively. In 1894 water was 

 brought into the barracks from a spring supposed to be pure. In 

 a short time eleven cases of typhoid fever declared themselves. 

 On examination it was discovered that the spring was contami- 

 nated, and the garrison ceased using it for a water supply. The 

 disease has now all but disappeared again. 



Similar facts in connection with typhoid fever have been 

 verified in more than twenty widely separated garrison towns. 

 Of the cases which still appear where filtered water is used, the 

 cause has invariably been found, when investigation was possible, 

 to be some accidental use of contaminated water. Thus at Nantes, 

 in Brittany, endemic typhoid fever was reduced by the use of 

 filters to isolated cases. In 1893 there were seventeen, and in 

 1894 there were thirty cases needing explanation. It was soon 

 remarked that nearly all were among orderlies, who have the 

 habit of taking their meals in certain restaurants of the town. 

 In each of these places it was discovered that the water used was 

 polluted by infiltrations from privies. In four other garrison, 

 towns the same fact was reported. 



A final instance, which is also one of the most remarkable, is 

 that of Auxerre. Here one hundred and twenty-nine soldiers 

 were down with typhoid fever in 1892. Filters were set up, and 

 there was one single case in each of the two following years. 



Only one objection has been urged against this triumph of the 



