[JOHNSTON-ELLSWORTH] ANNAHEIM METEORITE 79 



Method of Analysis 



The composition of the materials being unknown and the quanti- 

 ties small, it was necessary to use a scheme of analysis adapted to the 

 convenient identification and determination of as many elements as 

 possible on the one sample available. 



The material was dissolved in aqua regia at room temperature. 

 Sufficient sulphuric acid to displace the aqua regia and leave 0.5 to 

 1 c.c. excess was added and the solution evaporated on the water 

 bath. It was then heated to fuming, cooled, diluted, and examined 

 for silica. In the presence of silica the evaporation and fuming with 

 sulphuric acid was repeated. Copper was then precipitated by H2S 

 and weighed as oxide or metal. The solution after separating copper 

 was oxidized with bromine water, boiled with KOH in platinum and 

 filtered. The washed precipitate was ignited in a platinum crucible 

 and fused twice with sodium carbonate. The combined filtrates from 

 the treatment with KOH and the fusions were acidified with nitric 

 acid and P determined as magnesium phosphate or by the molybdate 

 method. 



Fusion with alkali carbonate was adopted as providing a con- 

 venient test for Cr, Mn, Al and V in addition to removing most of the 

 P. The H2O insoluble from the fusions was dissolved in HCl and 

 H2SO4 and Pt precipitated from sulphate solution. After removal of 

 the Pt, Ni and Co were separated by an initial precipitation of the iron 

 by ammonia followed by five basic acetate precipitations. The iron, 

 with the small amount of P remaining, was tested for Ti and finally 

 precipitated by ammonia and weighed as Fe203 + P205. The P2O5, 

 remaining with the iron, was subsequently recovered by the molybdate 

 method and the necessary corrections applied to the values for Fe 

 and P. 



In the above method no provision was made for the detection of 

 carbides, but the very minute amount of carbon found in analysis I 

 and its apparent occurrence as graphite locally disseminated in the 

 38% nickel alloy, as shown by metallographic examination indicated 

 that no important amount of carbide was to be expected. 



Finally, it should be noted, that the highest accuracy is not to 

 be expected in the results from the above procedure, especially where 

 only fractions of a per cent, of a constituent are involved, in the first 

 place because of the small weight of sample used, and secondly 

 because of the numerous operations required for the determination of 

 all constituents in one small sample. However, the work was care- 

 fully performed and the results are believed to be reliable. 



