370 T. HAGA 



adding a decinormal solution of iodine in potassium iodide to the 

 hydroxyamine solution in presence of alkali bicai'bonate, so long as 

 it is bleaclied, generally using starch as an indicator. In order to 

 get the best results, 1 find, firstly, that the solution for titration 

 should not be excessively dilute, and, secondly, that it should contain 

 very little alkali salts, and accordingly, neither contain, to begin 

 with, much acid, — for this needs to be neutralised, — nor receive dur- 

 ing titration more sodium bicarbonate than is necessary to take up the 

 hj'drogen iodide formed. The hydroxyammonium chloride used in 

 all my experiments had been prepared from mercury fulminate. 



Influence of the strcmjth of the solution :— - Very concentrated solu- 

 tions of hydroxyammonium chloride give results too low. For 

 example, 5 ces. of a solution of which the true titer was 10 ces. of 

 decinormal iodine solution, required, after evaporation to dryness and 

 addition of a few drops of sodium bicarbonate solution, 9.3 ces. only 

 of iodine ; 5 ces. of another solution with a titer of 9.9 ces. iodine 

 solution, required after evaporation only 9.4 ces., and 5 ces. of a third 

 solution with a titer of 9.55 ces. iodine required after evaporation 

 only 9.25 ces., the ditterences being not due to loss during evapora- 

 tion, but to irregular reaction with the iodine. Moderately con- 

 centrated solutions give good results, but when the dilution reaches 

 about 1000 to 1 the results become too high, and more and more so 

 with increase of the dilution. With a dilution of 3000 to 1 the 

 results of titration are very markedly too high, as may be seen from 

 the first of the tables which follow. 



With an iodine solution approximately decinormal and contain- 

 ing 0.1256 gram in 10 cub. centims., a solution of hydroxyammo- 

 nium chloride was used eontairiing 0.0330 gram in 10 ces., and 

 requiring by calculation 9.(5 ces. of iodine solution. Of this 10 ces. 

 were taken for each ex])eriment. 



