256 Note on Heusler 's Magnetic Alloy of , , and Copper .
By Andrew Gray , F.R.S. , Professor of Natural Philosophy in the University of Glasgow .
( Received December 15 , 1905 , \#151 ; Read January 25 , 1906 .
) In 1903 Fr. Heusler published the discovery of an alloy consisting of manganese , aluminium , and copper , which , in spite of the fact that it contained none of the so-called magnetic metals , iron , nickel , or cobalt , possessed striking magnetic properties .
Short accounts of work on the subject by Heusler and some other experimenters appeared , * but on the whole the discovery seems to have aroused comparatively little interest in this country before August , 1904 , when R. A. Hadfield exhibited a specimen of the alloy at the Cambridge meeting of the British Association .
At the beginning of the winter session 1904 an attempt was made to obtain some of this alloy for the Physical Laboratory of Glasgow University , with the view of determining magnetic curves for the material , and of otherwise extending our knowledge of this interesting manganese bronze .
The publication of a paper by Fleming and Hadfield f to a certain extent supplied the information sought , but also served to emphasise the fact that various samples of this alloy possess very different magnetic properties .
The maximum induction obtained by Fleming and Hadfield under a field of 200 C.G.S. units is less than one-half that obtained with a field of 150 in the case of one of Heusler 's numerous samples .
The form of the curves obtained by Fleming and Hadfield must also differ considerably from many of those indicated by the numbers given in Heusler 's papers .
Under these circumstances it has been decided to continue the work begun here , and to give now a short account of the work so far performed .
Dr. C. E. Fawsitt , of the Chemical Department of the University , became interested in the proposed research and attempted to make samples of the alloy .
The apparatus at his disposal , however , did not admit of production of the alloy in anything but small quantities .
The specimens obtained proved very retentive of magnetism , but were not suitable for the carrying out of quantitative work .
In the meantime two rods of the alloy had been obtained from Herr Heusler by Dr. G. E. Allan , one of the research students in the Physical Laboratory .
These rods were ground truly cylindrical on emery and * Cf .
1 Science Abstracts , ' Nos. 622 , 623 , 636 ( 1904 ) .
t ' Roy .
Soc. Proc. , ' June , 1905 .
Heusler 's Alloy of Manganese , Aluminium , and Copper .
257 tested by the magnetometric method .
The intensity of magnetisation ( I ) was calculated on the assumption that the effective lengths of the specimens were five-sixths of the actual lengths , and the magnetising field ( H ) was corrected hy using the demagnetising factors for cylindrical rods given by Du Bois .
The first of the specimens contained about 26'5 per cent , manganese , 14'6 per cent , aluminium , and the remainder copper .
This rod so far has only been tested in low magnetic fields ; the intensity of magnetisation induced by a field of about 8 C.G.S. units was approximately 105 .
The second rod obtained from Herr Heusler contained about 16 per cent , manganese , 8 per cent , aluminium , a little lead , and the remainder copper .
After having been dressed , this rod was found to be practically non-magnetic ( much less magnetic than is indicated by Curve 2 of the diagram ) .
As it was said to be from the same pouring as another which showed well-marked magnetic qualities , it was conjectured that the heating and vibration to which the rod had been subjected during the dressing operations had destroyed the magnetic quality , and an attempt was made to restore the magnetic properties by thermal treatment .
The rod was heated to 400 C. in Prof. A. Gray .
Note on Heusler 's Magnetic [ Dec. 15 , a furnace and allowed to cool slowly .
After having been placed in a magnetic field it was found to retain a considerable amount of magnetism .
It was next heated to 340 ' C. for about 20 minutes and allowed to cool , when it was found that the magnetic properties were much more pronounced .
With the specimen in this state it was put through a cycle of magnetisation , and the results are shown in Curve 1 of the diagram .
It was now decided to try the effect of extremely low temperature upon the material .
The specimen was immersed in liquid air , withdrawn , and put through a magnetic cycle as quickly as possible , the specimen warming up somewhat meanwhile .
The effect produced was extremely slight , but was towards an increase in magnetic susceptibility .
An endeavour was next made to get the specimen into a better magnetic condition by heating to various temperatures , * but no improvement was obtained .
Incidentally the critical temperature was found to be about 350 ' C. An attempt was now made to destroy the magnetic quality of the material .
Vigorous tapping at the temperature of the room was found to have no effect upon the residual magnetism .
Previous tapping at the temperature of 100 ' C. had been found to produce a considerable reduction in the residual magnetism , but the original value was restored by again applying the magnetic field .
It was thought that sudden cooling or " quenching " from above the critical temperature might permanently destroy the magnetic quality , and such was found to be the case .
The specimen was heated to 400 ' C. in the furnace and then plunged vertically into cold water .
Curve 2 of the diagram , which exhibits the results of a magnetic cycle carried through with the specimen after this treatment , shows the alloy to be in a comparatively non-magnetic condition .
An examination of the specimen showed it to have several cracks distributed over its surface as a result of the quenching , and this probably affected the magnetic tests to a certain extent .
The effect produced by the temperature of liquid air upon the material in its quenched and , at ordinary temperatures , nearly non-magnetic condition was now investigated and found to be very remarkable .
When tested at the temperature of liquid air the specimen was found to be more susceptible to magnetism than in its previous best condition , while it exhibited very much less hysteresis and retentiveness .
Curve 3 of the diagram illustrates the new magnetic condition and shows how , moreover , the comparatively high susceptibility thus given to the alloy disappeared as the temperature rose .
Curve 2 of the diagram was repeated after the temperature of the specimen had again become normal .
* Reference should be made to the extensive thermal experiments carried out by Heusler and his collaborators .
1905 .
] Alloy of Manganese , Alu , and Copper .
259 A microscopic examination of the material which we have been considering has been made .
After many trials it was found that a solution of ammonium hydrate formed an efficient etching agent ; the structure of the alloy , which had been revealed by polishing alone , was well brought out by this agent .
The work described here was carried out in the main by Dr. J. Muir and Mr. J. G. Gray , B.Sc. , two of my assistants .
We have now obtained a considerable quantity of the alloy in the form of rings and rods .
These specimens have been cast for us by Messrs. Steven and Struthers , the well-known Glasgow engineers and brass-founders , and some of them are being turned into elongated ellipsoids of revolution .
A continuation of the research on the lines indicated is in progress .
[ Addition , January 15 , 1906.\#151 ; Curves analogous to 1 , 2 , 3 of the diagram have been obtained from specimens of nickel and steel .
In the case of nickel the three curves differed only very slightly .
On the steel specimen quenching had , of course , a considerable hardening effect ; but the new magnetic condition was only slightly changed when the specimen was brought to the temperature of liquid air .
] VOL. LXXVII.\#151 ; A. u

