oO 



paper to call attention to two points in this process: First, though it has 

 been remarked by previous observers that there is a reaction between the 

 diver chloride formed and the ammonium sulphocyanide, is not this reaction 

 sufficient to produce an appreciable error? We here give some observa- 

 tions made, with this point in view. 



WITH DISTILLED WATER. 



These data seem to indicate that even though the time be reduced to a 

 minimum, the results are inconsistent and misleading. Now these varia- 

 tions may be eliminated by a process of filtering. Introduce a quantity of 

 sodium chloride, say 4cc from a ^\ normal solution, into a 200cc graduated 

 flask, add 4cc nitric acid (1.2 sp. gr.), free from nitrous acUl, and with dis- 

 tilled water at 15° C fill to mark. Mix well. When the silver chloride has 

 been separated, filter off lOOcc of the fluid through a dry filter. Introduce 

 the filtrate into a titrating flask, add 2— 3cc sat. sol. ferric alum and titrate 

 with the j\ nor. sol. am. sulphocyanide, till the addition of one drop causes a 

 light brown color to appear. This color once produced will be permanent. 

 The results of such a device are shown by the following data : 



These results agree with the quantities introduced and are constant. This 

 device was employed in estimating the chlorine in the waters from AVest 



