Smith — Determination of Chlorine in Natural \Yaters. 3G3 



lExample : 



■Second reading of tlie burette , 2 . 8 c. c. 



First reading of the bnrett^i ,. . . . 1 . 3 c. c. 



1 . 5 c. c. 

 Less error . .1 



i^o. c. c. silver solution required 1 . 4 c. c. 



Equivalent to chlorine 1.4m. 





The whole can then be readily calculated to parts per million 

 of water. In e:ood clear liaht where the conditions of the test 

 are kept perfectly uniform both in the comparison dishes and 

 the unknown solution the definiteness of the end-rea(?tion is sur- 

 prisingly clear. The accuracy of tliis method can be readily 

 proved by comparing results from definite quantities of the* 

 above salt solution and chlorine-free water. Silver solutions of 

 less strength can be used if desired where the amount of 

 chlorine is very low ; below one part per million, e. g., it may be 

 desirable to use a solution one-half or one-tenth as strong, i. e., 

 1 c, c. = 0.0005 or 0.0001 CI., but I have found that a solution 

 of a strength of 1 c. c. = 0.001 m. g. chlorine is usually the 

 l>est for all work. This method is to be recommended also be- 

 cause it does awav with the concentration of the waters, involv- 

 ing, as that does, loss from spattering and direct volatilisation of 

 the chlorides, and the necessary attendant loss of time. Where 

 waters parry extractive matters as do many of the river and 

 lake w^aters of this section of the country, receiving the lixivia- 

 tions from the swamps, imparting to them a high color, the 

 method, can be applied satisfactorily if some finely powdered, 

 absolutely chlorine-free alum, or better, a little precipitated 

 aluminum hydrate, be first added to precipitate the dissolved or- 

 ganic matters ; after standing over night the requisite amount 

 can be removed with a pipette. The color will then be dis- 

 charged, the water clear and bright with no masking of the 

 sharp definite end-reaction. 



When the chlorine is as high as twenty parts per million or 

 thereabout, it is advisable to add to the dish used for matching 

 the color enough standard sodimn chloride solution to represent 

 -about the same amount of chlorine as a preliminary test shall 



