80 PEOCEEDINGS OF THE AMERICAN ACADEMY 



the greenish, bhie-coloi'ed twine, was decomposed in that part of the 

 water naturally pure, while in the foul water the original color was 

 retained. This change was due to the local production of alkalinity 

 on that part of the twine in the purer water, as a consequence of its 

 polar state, resembling that of a metal wire. 



In an extreme case the copper wire, immersed in the silt at the 

 bottom of the water, actually presented scales of sulphide of copper, 

 formed after forty-eight hours' immersion, the temperature Q'2^° F. 



These points illustrate the generality of application in the method, 

 and lead me to notice particularly the chemical action on which it 

 depends, as a mistake has already been published. 



When in the decomposition of decaying matters the stage is reached 

 permitting the escape of sulphide of hydrogen, the ordinary tests of 

 lead, silver, and copper compounds detect the sulphur in the ordinary 

 simple way, or the odor of this gas indicates its presence. We have 

 entirely diffei'ent conditions under view when this test is applied. 

 The most sensitive test for free sulphur fails to detect its presence. A 

 compound of feeble stability, liable to change under the simple change 

 of temperature, or advance in fermentative action, is to be sought for 

 as a deleterious substance. The means used in ordinary cases fail 

 utterly ; but the electrical disturbance induced in a copper wire by a 

 simple difference of one part of its length immersed in one medium, 

 while the other part is wet by a slightly differing medium, is sufficient 

 to develop locally, — and locally only, — a change enabling us to obtain 

 evidence of the existence of the deleterious compound, by withdrawing 

 its sulphur, or other like element. The sulphur detected in this way 

 is presumably that which would act under natural conditions offensively 

 or noxiously. 



Since the varied application of this method to natural waters has 

 been extended, I have applied it successfully to detect sewage, when 

 mixed with or impregnating drainage water and well water. 



The method has been applied to the small volumes of water taken 

 in laboratory analyses, with the same good results. It is only neces- 

 sary to preserve a separation of the pure from the impure water, 

 through both of which the bright copper wire passes, the temperature 

 of 62° to 75° F. being kept constant. The impure water may be made 

 denser by common salt, so as to take the lowest place ; while in testing 

 the method I have made the impure water rest above a pure solution 

 of salt, with like results. 



The subject of the contamination of streams and other sources of 

 supply, on which whole populations depend, is becoming of grave im- 



