76 
ME. S. W. J. SMITH ON THE THERMOMAGNETIC ANALYSIS OF 
§ 16. In the third group of the table the percentage of nickel lies, in round numbers, 
between 35 and 40. 
A distinguishing feature of the analyses collected in this group is the length of 
time during which the solvent action of the acid upon the material was continued. 
In each case dilute HC1 was used of such strength (usually 1 of acid to 10 or 20 parts 
of water) that the solvent action upon the nickel-rich alloy (taenite) was very slow 
compared with that upon the nickel-poor alloy (kamacite). The acid was, in each 
case, frequently renewed, and the action extended over three months in three cases, 
and over two months in another. One analysis only (Canon Diablo, Florence) gave 
a less percentage of nickel than 34, viz. 31'54, and against this must he placed a 
subsequent analysis of a specimen of the same meteorite performed, purposely 
with the greatest care as a check upon the first, by Fahrenhorst and giving the 
value 35 - 66. 
It is important to notice that the total percentage of taenite in. the iron varied 
greatly from one meteorite to another. Thus, for example, Cohen obtained from about 
70 grammes of Bischtube iron enough taenite to enable him to use 0 - 4 gramme in 
his analysis; while out of 100 grammes of the Youndegin meteorite Fletcher 
obtained less than 0T gramme. The Glorieta meteorite was especially rich in taenite 
(containing apparently more than 10 per cent, of the latter) and was characterised by 
extremely thin alternating layers of kamacite and taenite. The removal of the 
kamacite (which was relatively very difficult) left a sponge-like skeleton of taenite, 
and during the action of the acid it was noticed that plates not more than 0T5 millim. 
thick sometimes consisted of alternating layers of kamacite and taenite. 
Under the circumstances of the analyses, as indicated above, the agreement amongst 
the numbers obtained by different observers is remarkable, and it seems difficult to 
avoid the inference that the ultimate residue, after the action of the acid, was in every 
case the same. 
But for the analysis of the Beaconsfield taenite, which is anomalous in more respects 
than its percentage of nickel (e.g., its specific gravity), the data so far considered 
agree with the view that the percentage of nickel in the nickel-rich constituent of 
taenite is not very different from 40. 
§ 17. The last point in connection with the chemical data, regarded in this way, is 
the information they afford concerning the percentage composition of the eutectic as 
a whole. 
If taenite is the eutectic, the action of the acid during long continued treatment of 
meteoric iron is twofold. In the first place it removes the layers of kamacite adjacent 
to the taenite, and in the second, when the action is sufficiently long continued, it 
removes the portion of the eutectic that consists of kamacite from the taenite itself; 
but if the taenite can be separated mechanically from the adjacent kamacite and 
analysed , the percentage of nickel found will be that contained by the eutectic. In the 
Welland and Cranbourne meteorites the mechanical isolation of the taenite was 
