116 M. Schneider on the Atomic Weights of Cobalt and Nickel. 



of a normal cliloridc is very difficult; and thirty-nine years ago, 

 when these determinations were made, the methods of preparing 

 pure cobalt and nickel compounds were very imperfect. 



Schneider's determination was made with the oxalates of 

 nickel and of cobalt, which have the advantage that they are 

 very insoluble in water and dilute acids, and also that the deter- 

 mination of the carbon and the metal may be made with great 

 accuracy. 



Great care was taken to obtain the oxalates quite pure. The 

 oxalate of nickel was prepared by adding to a feebly acid solution 

 of perfectly pure chloride of nickel, a cold saturated solution of 

 oxalic acid, washing the resulting oxalate with water until it was 

 perfectly neutral and free from all trace of acid, and then con- 

 tinuing the washing some time. When dry, it forms a sandy 

 powder, which under the microscope has a crystalline granular 

 appearance. 



The oxalate of cobalt was obtained by digesting pure carbonate 

 of cobalt with oxalic acid, and then washing out the oxalate, 

 at first by decantation, and then on a filter. When dry, the 

 salt is a loose woolly powder, which, like sulphur, becomes elec- 

 trical by rubbing in a mortar. 



For the analysis, the salts were heated to 110° C. in a stream 

 of di'y air. Tlie neutral oxalates contain for two equivalents of 

 carbon, one equivalent of metal ; and hence, from the quantita- 

 tive relation between the carbon and the metal, the equivalent of 

 the latter may be determined, and with the greater certainty from 

 the equivalent of carbon being accurately known. Schneider's 

 method consisted in the determination of this relation. It was 

 necessary that the oxalate should be perfectly neutral, and that 

 the salt used in each carbon and metal determination should 

 have the same degree of dryness. 



The determination of the carbon was made by an elementary 

 analysis, the usual precautious beiug observed. The metal was 

 determined by heating the oxalate in a bulb tube, in a stream of 

 atmospheric air, care being taken to prevent loss by spirting &c. 

 After the decomposition was over, the residue was heated in 

 a current of oxygen until every trace of carbon was burnt ; and 

 when this was effected, the oxide was reduced by hydrogen at as 

 low a temperature as possible, but the heat at length raised so that 

 the metal was superficially caked together. By this means the 

 metal bad completely lost the property of absorbing oxygen. 

 The hydrogen was then expelled by atmospheric air and the tube 

 weighed, the usual precautions being taken. The mean of the 

 determination of nickel gave as its equivalent 29'025 ; and of 

 cobalt, 30'003. The equivalents are hence simple multiples of 

 the equivalent of hydrogen ; the small difierence may be ascribed 



