98 MEMOIRS NATIONAL ACADEMY OF SCIENCES, VOL. XIII. 



Another fragment which gave no traces of carbon yielded the following: 



Fe Ni Si0 2 S 



1. 91.09 1.08 0.05 0.045 =92.265 



2. 92.08 7.05 =99.13 



Moissan also dissolved 53 kg. of Canyon Diablo in pure HC1. An insoluble residue weighing 

 800 grams was left, in which he found, besides carbon in the form of amorphous carbon, graphite 

 and diamond, an iron phosphide, P 2 Fe 3 , and silicon carbide (carborundum) SiC. The iron 

 phosphide occurred as little needles or as cubes. Analysis gave: 



Needles Cubes P 2 Fe 3 Calc. 



P 26.97 26.95 26.46 26.95 



Fe 71.63 72.15 73.51 72.43 73.05 



Ni trace. 



C trace. 



The silicon carbide occurred in little green hexagonal crystals. This was its first noted 

 occurrence in nature. 



Friedel 7 about the same time investigated the meteorite. He did not find the white 

 diamond in very large grains; but, in examining with the microscope the powder of the car- 

 bonado diamond, which was quite abundant, and taking care to immerse it in iodide of methylene, 

 he was able to discern among grams having exactly the appearance of small masses of the well 

 known carbonado, a certain number of small transparent grains which ho regarded as white 

 diamond. The statement of Moissan concerning the extreme heterogeneity of the Canyon 

 Diablo meteorite was confirmed by Friedel. He also mentions a bright compound, silver white 

 in appearance, fragile, less easily attacked by acids than iron, which occurred in comparatively 

 thick plates included between the crystals of nickel-iron, accompanied in this position by 

 laminae of schreibersite. This latter was distinguished by its ductility. These laminae stood 

 out slightly on a polished fragment. 



Friedel isolated a small quantity of this material and recognized it as a subsulphide of 

 iron containing 10.2 per cent of sulphur and 88.3 per cent of iron. There was also a small 

 quantity of phosphorus which could not be certainly measured on account of the small quan- 

 tity of material at disposal. The proportion above, if it could be considered as constant, would 

 correspond to the formula Fe 5 S. 



Besides this sulphide disseminated in the mass, there were nodules of yellow troilite, in 

 which, or in the neighborhood of which, carbon, that is to say, the mixture of ordinary carbon, 

 graphite, and diamond, appeared to be concentrated. The nodules were composed partly of 

 troilite, partly of carbon penetrated by this same sulphide. The nodules themselves were 

 enveloped in a thin layer of bright subsulphide. 



The occurrence of diamonds in the Canyon Diablo meteorites as well as others was dis- 

 cussed by Meunier 10 with the following conclusions: 



The irons in which diamonds are found are far from having the characteristics which may be called normal; they 

 depart greatly from the description which fits the typical meteoric irons. 



The latter, cut with the saw and polished, are of a more homogenous appearance than our finest steel, and appear- 

 ance which contrasts with their real heterogeneity; it is only under the action of acids that they show the network 

 of figures called Widmannstatten lines and which reveal the coexistence of alloys very unequally soluble. 



In the iron of Canon Diablo, it is enough to saw a surface, without polishing, in order to obtain, without the use 

 of an acid, a mosaic which may suggest the Widmannstatten figures, but which has nothing whatever in common with 

 them and which disappear upon polishing. These markings result from the existence, in the midst of the mass of a 

 nickel iron ore more or less homogenous, of laminae of a phosphuretted material known under the name of schreibersite, 

 occurring in small amount in ordinary irons where it is quite otherwise disposed. 



The iron of Magura (Arva), said to contain adamantine grains, belongs distinctly to the same lithologieal type as 

 that of Canon Diablo; the type under which comes also, with certain others, the Smith ville iron. These masses are 

 partly of the type designated by the name Arvaite in the collections of the French Museum. 



As to the irons which yield the beautiful Widmannstatten figures, they contain in general very little carbon and 

 sometimes not a trace of it. 



