24 
MR. S. W. J. SMITH ON THE THERMOMAGNETIC ANALYSIS OF 
The material is thus apparently remarkably free from elements other than those of 
the ferromagnetic group. 
After a ring about 2 centims. deep and 5|- centims. external diameter had been cut 
from the sample supplied, a solid disc-shaped core, about 3f centims. in diameter, 
remained. The ends and cylindrical surface of this were afterwards polished carefully 
and etched. At the conclusion of the experiments, the outer oxidised layer of the 
ring was removed on the lathe and the ring was then etched in order that its 
structure might be compared with that of the material which had not received 
thermal treatment. 
§ 3. Different samples of the turnings produced during the formation of the ring 
were kindly analysed by Mr. E. A. Wraight, of the Royal School of Mines, with the 
following mean result:— 
Fe = 92'84 (92'5 to 93'0 in different samples) 
Ni = 6-81 
Co = 0-37. 
In some of the samples there w r ere traces of chromium, Avhilst others contained up 
to 0 - 025 per cent, of sulphur. Examination of the polished surface of the core above 
mentioned showed the presence of an isolated speck of a brass-coloured mineral wdiich 
was almost certainly sulphide of iron (troilite), and it is probable that the sulphur 
found in some analyses was present in this form. No trace of carbon or phosphorus 
was found in the analyses. 
§ 4. Mr. A. Fowler, Assistant Professor of Physics in the Ptoyal College of Science, 
kindly examined the meteorite spectroscopically between the D lines and \ 4170 and 
found that there were no lines not accounted for by Fe, Ni, Co and Cr. The cobalt 
lines were relatively inconspicuous. Three chromium lines were identified with 
certainty. 
§ 5. The grosser structure of the meteorite was shown very clearly by the markings 
on the polished and etched cylindrical core. The distribution of the more and less 
nickeliferous layers was easily discernible. The actual 
thicknesses of the layers, and their geometrical arrange¬ 
ment with respect to one another, could be observed by 
tracing the course of individual layers across the plane and 
cylindrical surfaces of the core. The approximate relation 
between the positions of the layers observed is represented 
in fig. 1. 
All the layers observable were parallel to one or other 
of the four planes represented by their'traces upon the 
plane and cylindrical surfaces, viz., 00, 11, 22 and 33 
respectively. Some of the layers appeared to be appreciably 
curved. The numbers marked on the plane circular face in fig. 1 were judged to be 
the closest approximations that could be got to the general directions of the layers. 
