148 Proceedings of Indiana Academy of Science 



solution and suppresses iron, when present, by ammonium acetate and 

 tartaric acid. 



In this laboratory it was observed some years ago that when ethyl 

 alcohol is substituted for amyl alcohol and potassium thiocyanate for 

 the ammonium salt, a blue color is developed in the aqueous layer which 

 does not pass into the ether layer until mineral acid is added. It was 

 found that under these conditions cobalt could be readily detected in 

 nickel salts in which the procedure recommended by Treadwell failed 

 to show th? presence of this element. Of many salts examined only 

 one, a Kahlbaum's "kobalt-frei" NiSOj, failed to show the presence of 

 cobalt. Since nickel salts usually contain iron, a purple instead of blue 

 color appeared in the ether layer. It was found that shaking with a 

 few small crystals of sodium thiosulfate quickly reduced the iron and 

 a pure blue color remained. 



It appeared desirable to study this modification of the Vogel reaction 

 in greater detail in order to fix more certainly the conditions of con- 

 centration under which it is most delicate and reliable and to compare 

 the delicacy of the test in neutral and in acid solution. It seemed 

 probable that in acid solution the blue ether-.soluble compound might be 

 the acid of a complex anion, e.g., H::Co(SCN),, and that such an acid 

 might be isolated from the ether solution. Two such complex acids have 

 been obtained in crystalline form, H..Hg(SCN,)' and HAu(SCN) ,.2H,0.= 



EXPERIMENTAL. A solution of cobalt nitrate was prepared by 

 dissolving electrolytic nickel-free cobalt in nitric acid, removing excess 

 of acid by evaporation and diluting with water. This solution was 

 standardized (1) by igniting the nitrate and weighing as C03O,; (2) 

 by conversion to anhydrous sulfate. The cobalt content per cc. as found 

 by the first method was 8.401 mg. and by the second 8.391 mg. ; average, 

 8.396 mg. By accurate dilution solutions were prepared containing 

 respectively, 1.0, 0.1, 0.01 and 0.001 mg. of cobalt per cc. Standard 

 solutions of ammonium, potassium and sodium thiocyanates and of 

 hydrochloric and sulfuric acids were also prepared. 



A. Treadicell-Vogel Test (Neutral solution). The eff'ect of vary- 

 ing the concentrations of the various components of the test solution 

 was studied and the conditions determined under which the test is 

 most delicate. The optimum conditions were found to be high concentra- 

 tion of thiocyanate, small total volume and small volume of the ether- 

 alcohol layer. For example, 5cc. of 4 normal Na, K, or NH, thiocyanate, 

 2.5 cc. of saturated ammonium acetate and 0.-5 cc. of cobalt nitrate 

 (0.001 mg. per cc. Co) when shaken with 1 cc. of 1:1 mixture of amyl 

 alcohol and ethyl ether gave a faint but distinct blue color in the ether- 

 alcohol layer when this was observed against a white background. Under 

 these conditions 0.0005 mg. of Co in a total aqueous volume of 8 cc. is 

 clearly detectable. Treadwell's statement that 0.02 mg. of cobalt is 

 detectable is conservative. Not much difference was observed in the 

 effectiveness of the three alkali thiocyanates but it was noticed that 



' Rosenheim ami Cohn, loc. cit. 



= Bjerrum and Kirschncr, Die Rhodanide des Goldes Mem. acad. roy. sci. lettrcs 

 Dancmark, Smc seric, V, No. 1, p. 20. 



