EXAMINATION OF WATER. 427 



Chemically speaking*, it is a good sample because the solu- 

 ble constituents in one drop are present in relatively 

 the same amounts as in a liter or barrel of the water.. 

 This is not true with reference to the suspended bacteria. 

 These are solid particles, and as such tend to settle or 

 subside. The so-called water purification in many instances 

 may largely consist of mechanical sedimentation. Al- 

 though, therefore, the mass of the water may still contain 

 some typhoid bacilli, the small quantity taken for examin- 

 ation (1-10-20 drops), may, or may not, contain one or more 

 of these organisms. The few specific germs are easily 

 masked by the hundreds and thousands of common water 

 bacteria that are usually present. If necessary, the num- 

 ber of organisms taken for an examination may be increased 

 by prolonged centrifugation of the water. Moreover, the 

 separation may be favored by the addition of typhoid serum 

 or of histon. 



The typhoid-like or pseudo-typhoid bacilli that may be 

 present require careful and prolonged study. Recent inves- 

 tigations have shown that typhoid bacilli, or at all events 

 organisms that cannot be distinguished from these by any 

 known test, may be present in water, in soil; also in the feces 

 of healthy persons who have never had typhoid fever. 

 Thus, Losener isolated 5 such organisms from tap-water, 

 field soil, normal pus, from a hog cadaver buried 4 weeks, 

 and from a typhoid spleen buried 96 days. Pfeiffer and 

 Kolle have pronounced these organisms to correspond ex- 

 actly to the typhoid bacillus, even to the application of 

 Pfeiffer's reaction, and hence their typhoid character cannot 

 be questioned. Remlinger and Schneider, employing Eis- 

 ner's medium and working under the direction of Vaillard 

 at the Val-de-Grace hospital at Paris, isolated a bacillus in- 

 distinguishable from that of Eberth: 



8 times out of 36 samples of water; 



6 times out of 10 samples of soil; 



3 times out of 8 samples of normal feces. 



