METEORIC AND ARTIFICIAL NICKEL-IRON ALLOYS. 
75 
§ 14. The earlier data contained in the above table are doubtful, since they represent 
the results of the first attempts at analysis of taenite. For instance, in some of 
Meunier’s experiments the kamacite and taenite were separated by means of the 
difference in colour, blue in the one and yellow in the other, produced by superficial 
oxidation of a powdered sample of the iron. Since careful microscopic examination 
shows that taenite and kamacite can occur in alternating layers, each not more than 
a small fraction of a millimetre in thickness, it is clear that such a method of 
separation cannot be relied upon. The difficulty of completely separating nickel and 
cobalt from iron by chemical analysis has, moreover, been realised more fully since the 
date of these experiments. 
Of the two later analyses giving a percentage of nickel in taenite less than 20, it is 
not claimed for the second that it is more than “fairly correct.” Only 0‘02 gramme 
of material was used in the analysis, and this amount was obtained by a laborious 
mechanical selection from the rest of the iron. 
With respect to the analysis by Tassin, it is not certain from the description how 
the sejiaration was effected. 
A later analysis by Tassin of a sample of taenite from another meteorite yielded 
36 per cent, of nickel (‘ U.S. National Museum,’ vol. 28, p. 215, 1905). 
§ 15. In the second group of the table are collected the analyses which give values 
for the nickel (together with cobalt) content ranging between 25 per cent, and 
30 per cent. 
In three at least of the seven analyses (those by Davison and Flight respectively) 
the method of separation was mechanical. No chemical solvent was used. In the 
paper describing the analysis of the Cosby Creek (Sevier) meteorite no details are 
given, but the material was apparently picked out of portions of the iron which had 
been disintegrated by oxidation. It is significant to note that this taenite and that 
analysed earlier by Reichenbach (and giving 13'56 per cent, of Ni) were obtained 
from different specimens of the same meteoric iron (see Cohen, I., p. 103). Of the 
two analyses given by Davison, one refers to the comparatively broad bands 
of taenite (thickness -jA to - 3 j ( y millim.), and the second to the narrower bands 
(thickness about T oo to ^oo millim.) picked from the so-called “plessitic” layers of 
the iron. 
With respect to the specimens of taenite in this group, which were separated 
chemically by the continued action of dilute hydrochloric acid, it is noteworthy in 
connection with the interpretation which follows below, (1) that the physical properties 
of the Magura taenite agree closely with those of the plates picked by Flight from the 
Cranbourne meteorite, and that the iron is described by Weinschenk as soluble with 
difficulty ; (2) that the plates of Staunton taenite were relatively thick and rich in 
carbon ; (3) that the physical properties of the Misteca taenite were intermediate 
between those of the Staunton taenite and those of several almost carbon-free taenites 
referred to below. 
l 2 
