234 C. F. Roberts — Estimation of Chlorates and Nitrates. 



oxide left in the Hempel burette. The results varied from one 

 another by 4 or 5 per cent on different portions, but a few 

 trials were sufficient to indicate that the trouble lay in the sil- 

 ver nitrite and not in the process itself. The percentage of 

 purity as obtained from several trials by ignition was 98*4 per 

 cent. The average of ten trials estimating from nitric oxide 

 gave 98 '1 per cent, and from the liberated iodine 98*7 per 

 cent. This was uniformity enough to indicate that the method 

 of work was trustworthy, although there was apparently such 

 a lack of homogeneity in the silver nitrite as to make it impos- 

 sible to use it in testing the value of the formula given above. 

 To obviate this difficulty, more silver nitrite was precipi- 

 tated, well washed, dissolved in warm water, and a solution of 

 sodium chloride added. The silver chloride was then filtered 

 off, and the resulting solution of sodium nitrite was used in 

 the following experiments. To determine the strength of this 

 solution, ten or fifteen cubic centimeters were drawn off from 

 a burette, and the nitrite estimated with potassium permanga- 

 nate according to the method of Kinnicutt and Nef.* At the 

 same time, similar amounts were introduced into the retort 

 with manganous chloride, and the nitrite estimated both from 

 the liberated iodine and the nitric oxide formed. It was found 

 that the air must be thoroughly driven out of the apparatus 

 before the nitrite was introduced, as the carbon dioxide, pass- 

 ing over the solution, decomposed it with liberation of nitric 

 oxide which, in presence of air, gave an excess of free iodine, 

 and at the same time a little gas was lost. Accordingly, car- 

 bon dioxide was first passed through the apparatus for some 

 time, then the nitrite was introduced through the funnel-tube 

 and rinsed in with a little water, followed by the manganous 

 chloride solution, care being taken that the water should not 

 exceed one-third of the total volume of the liquid, according 

 to the precaution to be shown necessary by Gooch and Gruener. 

 Working in this way with the solution of sodium nitrite, the 

 following results were obtained : 





Vol. 



Weight reckoned 



Weight reckoned 



Weight reckoned 





taken. 



from KMn0 4 . 



from NO. 



from iodine. 





cm. 



grm. 



grm. 



grm. 



(1) 



10 



0-0463 



0-0456 



0-0450 



(2) 



10 



00460 



0-0460 



0-0470 



(3) 



15 



0-0704 



0-0708 



0-0722 



(4) 

 (5) 



15 



0-0690 



0-0712 



00750 



15 



0-0701 



0-0704 



0-0722 



(6) 



15 



0-0688 



0-0696 



0-0695 



(?) 



15 



0-0706 







(8) 



15 



0-0712 







* Ann. Chem. Journ., v, p. 388. 



