88 
Proceedings of the Royal Society of Edinburgh. [Sess. 
XII. — Heusler’s Magnetic Alloy. By Alexander D. Ross, M.A., B.Sc., 
Houldsworth Research Scholar, University of Glasgow. Communi- 
cated by Professor A. Gray, F.R.S. 
(MS. received March 26, 1907. Read May 6, 1907.) 
Since the discovery by Heusler in 1903 of a magnetic alloy of copper, 
manganese, and aluminium, considerable interest has been aroused by the 
peculiarities of its properties and the difficulty of their explanation. In 
the summer of 1905 a preliminary investigation of a sample of the alloy 
was made in the Physical Laboratory of Glasgow University, and the 
results obtained have been published in a paper by Professor A. Gray.* 
The tests now described were carried out on an alloy with a somewhat 
lower percentage of copper, the composition being: — 25 per cent, man- 
ganese, 125 per cent, aluminium, a trace of lead, and the remainder copper. 
The specimens were cast in the form of rings and rods by Messrs Steven 
& Struthers of Glasgow. So far only the rods have been used, the tests 
being made by the magnetometric method. In order to avoid any difficulty 
regarding the effective lengths of the specimens, elongated ellipsoids of 
revolution have been used instead of cylindrical rods. The factors in- 
vestigated by Ewing have been employed to give the demagnetising force 
due to the specimen and the true magnetising field (H). 
With a view to testing the apparatus, preliminary I-H cycles were 
taken with a rod of the alloy about 60 cms. long and P30 cms. in diameter. 
It was tested in the condition as supplied by the makers, the casting being 
simply “ dressed,” and the ends cut square. A portion about one-third the 
length of the original rod was cut off and retested. On applying the 
corrections investigated by Du Bois, the two curves were in good agree- 
ment. The alloy exhibited very little hysteresis, and the saturation point 
was well marked. 
A specimen, which had been ground on emery to an ellipsoidal form, 
was subjected to thermal treatment. After being taken through a cycle 
(fig. 1, curve 1) it was heated for ten minutes to 50° C., slowly cooled, and re- 
tested. No change in magnetic quality was observed. Successive heatings 
were then made to 100° for a few minutes, to 100° for two and a half hours, 
and to 145° for a few minutes. In each case the specimen was slowly cooled 
* Proc. Roy. Soc., Sect. A, 77, p. 256. 
