166 



time required for the pink precipitate to appear was not recorded. The 

 samples were observed after standing '2\ hours. Prom the results tabulated 



below it is apparent thai the tesl is at least as delicate in the presence 

 in the absence of cyanide and that the results are obtainable much more 

 quickly from the complex than from the simple ion. In the extreme dilutions 

 of the simple ion the precipitate was frequently a single r< d crystal very 

 minute and difficult to see. 



TABLE I. 





T 



I M K . 







Molarity. 







Mg. Ni per cc. Ra 



iO Ni : H .<) 





NISOj 



K 2 Ni(CN)< 





. 0005 



Immediate 



Immediate 



(i 02934 1 



34.000 



. 00005 



1 hour 



3 min. 



. 002934 1 



340.00(1 



.00001 



24 hours 



5 min. 



.000587 1 



1 .700.000 



. 000009 



24 hours 



10 min. 



. 0005 2 S 1 



1 .900.000 



000008 



24 hours 



20 min. 



.000470 1 



2.130.000 



000007 



24 hours 



30 min. 



.000411 1 



2,430,000 



000006 



24 hours 



1 hour 



. 000352 1 



2,840,000 



000005 



24 hours 



21 hours 



000293 1 



3.400.000 



. 000004 



No ppt. 



2-1 hours 



. 000235 1 



4 ,260,000 



000003 



No ppt. 



Xo pink eolor 



.00017H 1 



5.700,000 



. 000002 



No ppt. 



No pink color 



.000117 





3. Oxidation of Cobaltocyaxide Iox to Cobalticyaxide Iox. 



When KCX is added to a solution of cobalt salt, brown-red Co(CN)s 



is first precipitated and then redissolved to a brown solution of K-jCoiCX I . 

 On heating this soon changes to a pale yellow and the color change is generally 

 assumed in manuals of analysis to indicate the completion of oxidation to 

 eobaltieyanide. We at first proceeded upon this assumption, but when the 

 first drops of AgX0 3 were added to some of our complex cyanide solutions, 

 soon after the eolor change took place, the solution darkened and addition 

 of more AgX0 3 produced a dark-gray precipitate while solutions which had 

 stood for several hours did not darken and gave a pure white precipitate. 

 When one of the darkened solutions became distinctly opalescent, we sus- 

 pected that colloidal silver had been formed. This was explainable by the 

 assumption that AgX0 3 had been reduced by cobaltocyanide which was still 

 present according to K 4 Co(CX)r. + AgX0 3 = K 3 Co(CX),- -f Ag + KX0 3 . 



