7
UNIVERSITY OF TORONTO
STUDIES
PAPERS FROM THE PHYSICAL
LABORATORIES
UNIV. OF MAIGA.
MAY 7 1007

No. 20: ON THE MAGNETIC PROPERTIES OF HEUSLER'S
ALLOYS, BY J. C. MCLENNAN.
(Reprinted FROM THE PHYSICAL REVIEW, VOL. XXIV.)
THE UNIVERSITY LIBRARY: PUBLISHED BY
THE LIBRARIAN, 1907
University of Toronto Studies
COMMITTEE OF MANAGEMENT.
Chairman: ROBERT ALEXANDER FALCONER, M.A., Litt. D., LL.D., D.D.
President of the University
PROFESSOR W. J. ALEXANDER, PH. D.
Professor W. H. ELLIS, M.A., M.B.
Professor A. KIRSCHMANN, PH.D.
PROFESSOR J. J. MACKENZie, B.A.
PROFESSOR R. RAMSAY WRIGHT, M. A., B.Sc., LL. D.
Professor GEORGE M. WRONG, M.A.
General Editor: H. H. LANGTON, M.A.,
Librarian of the University
!
S
७ ५.
[Reprinted from the PHYSICAL REVIEW, Vol. XXIV., No. 6, June, 1907.]
*
ON THE MAGNETIC PROPERTIES OF HEUSLER'S
ALLOYS.
SINCE
By J. C. MCLENNAN.
INCE the discovery by Heusler some three years ago that it
was possible to prepare alloys of copper, manganese, and alu-
minium which possessed magnetic properties similar to and com-
parable in amount with those of cast iron, considerable attention
has been paid by a number of investigators' to this interesting phe-
nomenon.
Among others Mr. H. F. Dawes, M.A., and Miss L. B. Johnson,
M.A., demonstrators in the Physical Laboratory at Toronto, and
Messrs. H. A. McTaggart and J. K. Robertson, senior students in
the same laboratory have from time to time during the past year,
as opportunity offered, studied the properties of some of these al-
loys which they prepared, and the following paper contains a few
notes on the results of their work, which is still being continued
by two of them.
In order to prepare samples of the alloys a supply of manganese,
copper, and aluminium, sold as chemically pure and free from iron
was obtained from Messrs. Eimer and Amend of New York.
The
¹Fr. Heusler, W. Starck and E. Haupt, Verh. der Phys. Ges., 5, p. 219, 1903.
Heusler, Starck and Haupt über der ferromagnetischen Aigenschaften von Legierung
unmagnetischer Metalle, Marburg, 1904. Gumlick, Ann der Physik, 16, p. 535, 1905.
Austin, Verh. der Deut. Phys. Gesell., 6, p. 211, 1904. Fleming and Hadfield, Proc.
Roy. Soc., 76, p. 271, 1905. E. Take, Verh. der Deut. Phys. Gesell., 7, p. 133, 1905,
Ann. der Physik, Vierte Folge, Band 20, p. 249. Hill, PHYS. REV., 21, p. 335,
1905; 23, p. 498, 1906. Guthe and Austin, Bulletin of the Bureau of Standards,
Washington, Vol. 2, No. 2, page 297, Aug., 1906.
449
450
[VOL. XXIV.
J. C. MCLENNAN.
metals were melted in an ordinary coke furnace, such as is used in
brass foundries, and poured into moulds just as in making brass
castings. A graphite crucible was used for the purpose, and in the
process the copper was first heated until it reached a molten state,
after which the manganese was added to it. When this in turn was
melted the aluminium was added, and the molten mass then stirred
for a short time with a porcelain rod, after which the crucible was
taken from the furnace and the melted metal poured into the moulds.
For a preliminary set of the alloys four castings were made. As
Heusler had found that the most highly magnetic alloy contained
manganese and aluminium in the proportion of their atomic weights,
one casting was made with as nearly as possible this proportion of
the two metals. A second was made with a larger proportion of
aluminium and two others with smaller proportions of this
constituent.
The percentages of the metals used in each of the four castings
and the proportions of aluminium to manganese by weight and by
atomic ratios are shown in Table I.
TABLE I.
Composition of Alloys - Percentages.
Number of
Alloy.
Ratio of
Manganese to
Aluminium
Aluminium.
Manganese.
Copper.
Atomic Ratio
of
Aluminium
by Weight. to Manganese.
I.
8.0
32.1
59.8
.250
.51
II.
9.7
25.6
64.6
.378
.77
III.
14.3
28.6
57.1
.498
1.01
IV.
15.9
23.9
60.3
.683
1.392
In each casting sufficient material was used to make two cylin-
drical rods about 20 cm. long, and 8 mm. in diameter, as well as
two rings about 7 cm. in diameter, and 2.5 cm. in height. It was
thought desirable to make the castings in the form of rings as well
as rods, in order to investigate more fully the permeability of each
specimen.
In the whole of the operation the greatest care was taken to keep
the alloy free from iron in order that there might be none of this
metal present to vitiate the results or to cast doubt upon their
validity.
No. 6.] MAGNETIC PROPERTIES OF HEUSLER'S ALLOYS.
45I
The investigations described below include one by Mr. Dawes on
the phenomenon of magnetostriction exhibited by one set of the
rods, one by Messrs. McTaggart and Robertson on the same phe-
nomenon with the second set of rods, and one by Miss Johnson on
the permeability of the alloys cast in the form of rings.
1
LG
BA
A. MAGNETOSTRICTION. (EXPERIMENTS BY MR. DAWES.)
I. Apparatus. The apparatus used in measuring the elon-
gation was, with some changes, modelled after that used by
Rhoads in his investigations on magnetostriction in iron. The
apparatus is shown in section in the accompanying diagram, Fig. 1.
A long hollow solenoid L produced the magnetic field in which
the rod to be tested was placed. The
lower end of the rod F was screwed
into the fixed brass stand H, and the →
upper end carried a brass rod E. A
cylindrical water jacket K placed
between the coil and the rod served
to keep the latter at a constant tem-
perature. The change of length of
the rod was magnified by means of
an optical lever. A brass block D
carrying a brass pivot indicated by
the inverted arrow was clamped to
the rod E, and so moved up or
down as the rod F was elongated or
shortened. A brass tube G which
was attached to H at the base of
F supported a horizontal projection

K
Fig
Fig. 1.
A. A carried two brass pivots indicated by the second arrow.
A lever B rested horizontally between these sets of pivots as
indicated. At C a mirror was mounted so that it could rotate
about an axis perpendicular to the plane of the paper and attached
to A. This mirror was supported in a vertical plane by an attach-
ment resting on the end of B. The brass points rested on agate beds
and the axis was jewelled. The image of a scale as seen in C was
¹ Rhoads, PHYS. REV., Vol. VII., No. 2, p. 65, 1898.
452
[VOL. XXIV.
J. C. MCLENNAN.
observed by means of a telescope. In order to trace the method
by which the change of length was measured, suppose that the rod
was slightly elongated relative to the tube G. The brass point
carried by D, would then move upwards, and in consequence the
C-end of the lever B would be allowed to drop. This would.
cause the mirror to be tilted backwards, and hence cause a change
in the scale reading as observed by the telescope. It will be seen
that if the change of length was very small, it would be propor-
tional to the change of scale reading, and the proportionality factor
could be found by measuring the ratio of the arms of the lever, the
distance between the axis of rotation of the mirror and its point of
support, and the distance of the scale from the mirror.
In the present experiment a modification was made in the ap-
paratus as used by Rhoads. Formerly a plane mirror was used at
C and the scale was placed at a distance of about 4 meters from it.
This mirror was replaced in these experiments by a concave mirror
of one meter radius of curvature, and the scale was placed at a dis-
tance from it of one half meter, the focal length of the mirror. By
this means the image of the scale was magnified sixteen times its
former size and the sensibility, i. e., the change of reading for a given
change of length was double that obtained with the plane mirror.
A glass scale was used in the telescope instead of a cross hair, and
the number of divisions on this scale over which a line of the image
moved was taken as the reading. The sensibility of the arrange-
ment was such that a change of reading of one division resulted
from a change in length of the specimen of 10-5 mnm.
II. Measurements. In making measurements the change of
scale reading was found for a series of applied fields ranging from about
30 to about 900 C.G.S. units. The strength of the applied field
was calculated from the value of the current, and the known dimen-
sions of the coil and the change of length per centimeter was de-
duced from the scale readings. For each series of readings curves.
were plotted, the strengths of field, H, being taken as abscissæ, and
the changes per unit length multiplied by 107 as ordinates.
In the first few series of measurements, the readings were taken
by applying each field for an instant and then withdrawing it.
By this procedure, an average result was obtained for the effect
No. 6.]
MAGNETIC PROPERTIES OF HEUSLER'S ALLOYS.
453
investigated, but the method was not suitable for bringing out any
hysteresis effect which might be present. It was therefore dis-
carded, and the method of increasing the field step by step and
then reducing it in the same way was adopted.
It was found that a cycle of readings could be taken in this way
in about two minutes.
When the current was kept passing through the coil it might be
expected that the heating would effect the change of length. In order
to test for errors from this cause, a brass rod was inserted in the
apparatus in place of one of the specimens investigated, and a cycle
of fields similar to that used in working with the alloys applied to
it. As brass is non-magnetic and consequently does not exhibit
the phenomenon of magnetostriction, the development of heat by
the current and its diffusion into the walls of the water jacket would
be followed by the elongation of G, Fig. 1, which would produce
an apparent shortening of D. This test showed that during the time
of a cycle of observations no error from heating was introduced.
Elongation curves for each of the four specimens are shown in
Figs. 2-5, the corresponding sets of values being given in Table
Curves 1 and 3 A were taken by the first method, the others
by the method of cycles.
II.
TABLE II.
Alloy No. 1.
dl
H
ΙΟΥ
dl
H
•
ΙΟΥ
33
0
50
0
220
900
ir is
.3
.5
130
.1
Alloy No. 2.
33
0
495
- .5
95
.15
375
— .75
225
.25
335
-1.0
335
.25
286
-1.0
375
.25
225
-1.25
491
-.25
95
-1.5
-.5
945
-.25
454
[VOL. XXIV.
J. C. MCLENNAN.
Alloy No. 3.
Curve A.
67
2÷5
22
.25
315
3.8
44
.5
352
4
.6
385
4.2
90
.8
435
4.5
124
1.0
456
5.5
192
2.0
980
9
255
2.3
276
2.5
Alloy No. 3.
Curve B.
3333
.5
470
32
56
1.4
385
31
79
4.4
351
30.4
100
8
310
30
143
15
260
29
215
20
215
27.5
260
23.2
143
22.8
310
26
100
19
351
27
79
14
385
27.4
56
9.5
470
27.5
33
5.8
910
33
0
3
Alloy No. 4.
Curve A.
33
.5
912
3.2
56
1
475
3
85
1.5
375
3
110
143
220
275
320
385
475
~ ci ci
357
2.5
2.4
320
2.3
2.5
275
1.5
2.5
221
1
si si
2.5
143
.5
2.5
110
0
2.6
85
-2.5
Alloy No. 4. Curve B.
3333
2
490
11.5
56
5.5
400
11.5
90
7.25
332
11.5
111
8.5
221
11
170
9.5
170
11
7
221
11
110
11
332
11.5
90
10.5
480
11.5
56
10
940
23
33
16
0
3
1
No. 6.] MAGNETIC PROPERTIES OF HEUSLER'S ALLOYS.
455
III. Discussion. — A careful examination of these and the other
curves found in the series of measurements showed that three fac-
tors entered into the change of length of the bar.
1. When a change was made in the field to which the bar was
subjected an almost instantaneous change was made in the length.
This change consisted in a lengthening as the field was increased,
and a shortening as it was decreased.
2. As the field was decreased there was a certain lag or hystere-
sis effect so that the length did not have the same value for a given
field as the field was approached from above or below.
3. In some cases when the specimen was left in a field for some
time there was a continuous shortening. This effect when it exis-
ted was very marked in strong fields so that the time for which the
specimen was left in these fields had a comparatively great influence
on the length. This effect was also observed by Austin in some
experiments made by him with these alloys. It does not appear,
however, to have been found by him in some measurements he has
recently made, in collaboration with Guthe, with other samples of
the alloys.
The form of the curve arising from a cycle of readings on a given
specimen would depend, therefore, on the relative importance of
x10?

45
이
​100 200
300 400 500
Fig. 2.
600
700
800
900
H
these three factors. In the early stages of an increasing field there
would be no effect from (2) and that from (3) would be small so
that the initial portions of the elongation curve would correspond to
In each of the
the change of length arising from the first source.
curves shown, this part of the curve exhibits the same character.
The curve rises slowly in the very weak fields and then in fields
extending from about 50 to 200 units quite rapidly. In still higher
fields the rise becomes smaller and tends to a maximum value.
In the case of alloy No. 2, Fig. 3, it will be seen that while the
456
[VOL. XXIV.
J. C. McLENNAN.
actual changes were small compared with those in alloys No. 3 and
No. 4, the third factor was of more importance than the first, with

S
0
-5
100 200 300 · 400 500
600 700 800 900 H
Fig. 3.
the result that in the higher fields the bar diminished a little below
its original length. This was also observed in the case of alloy No.
I, although it is not shown in the curve as the experimenter failed to
keep a record of his measurements.
In curves 3, B and 4, B, Figs. 4 and 5, this cause of change of
dl~10?

20
15
10
A
B
100
200 300 400 500 600 700 800 900 H
Fig. 4.
length was not sufficiently great to make the return half of the curve
cut the forward half, i. e., to overcome the hysteresis effect. But if
the strong fields had been left on a little longer this would probably
dl.107

15
10
B
A
100
200 300 400
500 600
700 800 900
H
Fig. 5.
have been the case and there would have been a final resultant con-
traction. An example of this effect is shown in 4, A, Fig. 5, where
No. 6.] MAGNETIC PROPERTIES OF HEUSLER'S ALLOYS.
457
the return portion of the curve crosses the forward initial portion
and consequently masks the presence of the hysteresis effect shown
in 4, B.
In Figs 4 and 5, curves A represent the elongation obtained with
two of the alloys when magnetized for the first time, and curves ы,
in the same figures represent the elongations obtained with the same
rods after they had been magnetized and demagnetized several times,
and then allowed to lie undisturbed for about thirty hours.
This increase in the elongation has not hitherto been observed,
and it may possibly be connected with the increase of permeability
observed by Starck and Haupt,' Gumlich,2 Hill 3 and others with
these alloys when they were heated for some time to a moder-
ate temperature, such as a 110° C. after being cast.
It may pos-
sibly, too, point to changes in the constitution of the alloy analo-
gous to those marking the recovery of iron from overstrain observed
by Muir,* on heating strained iron bars for short periods in boiling
water.
Considering the curves as a whole it will be seen that the elonga-
tion increased with the percentage of aluminium, present in the alloy,
the maximum elongation, curve B, Fig. 4 being obtained with the
alloy in which the manganese and aluminium were present in the
proportions of their atomic weights.
The maximum elongation in this case, it will be seen, was 16 ten-
millionths or about one third the maximum for iron.
The effect of higher percentages of aluminium is shown in Fig. 5,
curve B, where the maximum elongation was 11.75 ten-millionths.
Chemical analyses of the alloys have not as yet been made, but
if the constituents of the different samples of the alloys can be taken
to have the same relative proportion as the metals from which they
were cast, these results would go to show that the maximum elonga-
tion was obtained with alloys containing the same proportions of
manganese and aluminium as were found by Heusler to give the
highest permeability.
¹ Deutsch. Phys. Gesell. Verh., 5, 12, pp. 224–232, June, 1903.
2 Electrotech. Zeit., XXVI., 9, Mar., 1905.
3 Hill, PHYS. REV., Vol. XXIII., No. 6, p. 498, 1906.
+ Phil. Trans. Roy. Soc., London, Vol. 193, pp. 1-46, Series A.
·
458
[VOL. XXIV.
J. C. MCLENNAN.
B. MAGNETOSTRICTION. (EXPERIMENTS BY MESSRS. H. A.
MCTAGGART AND J. K. ROBERTSON.)
In the experiments made with the first set of rods a very marked
difference was observed in the elongations obtained with alloys Nos.
3 and 4, when magnetized for the first time, from what was obtained
with them after they had been magnetized and demagnetized several
times. This result pointed to the existence of an unstable equili-
brium among the molecules of newly cast specimens of the alloys,
and it forms a corroboration of the opinions of other experimenters
who have arrived at the same conclusions from their observations
on the changes in permeability of the alloys produced by variations.
in temperature.
TABLE III.
Alloy No. 2.
dl
H
2.107
dl
ΙΟΥ
H
ΙΟΥ
Τ
235
0.8
300
-0.8
370
1.7
218
-1.7
531
2.5
155
-2.5
675
2.5
0
-3.3
440
0
Alloy No. 3.
55
1.6
580
13.3
96
3.3
376
12.5
130
4.2
255
1
10.8
198
6.6
202
10
255
8.3
155
8.3
272
9.2
118
6.6
376
10.8
75
4.2
474
12.5
0
-1.6
Alloy No 4.
32
4.2
255
15.8
62
8.3
180
15
96
10.8
125
14.2
145
12.5
84
12.5
202
14.2
55
10
250
15
40
8.3
355
15.8
10
4.2
474
16.7
0
0
595
17.1
No. 6.] MAGNETIC PROPERTIES OF HEUSLER'S ALLOYS. 459
It seemed desirable to investigate this point more closely, and
therefore a set of measurements was made by the above mentioned
senior students on the elongations of the second set of rods cast at
the same time as those used by Mr. Dawes. Before proceeding
with this however, a set of readings was taken with the rods used by
the latter to see whether any change had taken place in them in the
six months interval during which they had been laid aside. The
results of their measurements are given in Table III. and represented
graphically by the curves in Figs. 6–8.
+5
With alloy No. I no elongation whatever could be observed by
them. With alloy No. 2,
Fig. 6, the general form of
the curve they obtained is
similar to that given in Fig.
3, by Mr. Dawes, although
the maximum elongation

they obtained in increasing
100
200
300 400
500 600 H
Fig. 6.
fields and the maximum shortening in decreasing fields was greater
than that observed by him with this specimen.
Their results with alloy No. 3, Fig. 7 show the presence of the
three effects pointed out in the first set of measurements, the
hysteresis and the shortening being clearly marked. The maximum
272107


15
24x10
10
5
-51
100 200 300 400
500
Fig. 7.
15
10
S
100
200 ૩૦૦ 400
⚫500
Fig. 8.
elongation obtained by them with this specimen corresponded to a
field of 600 units and is slightly less than that obtained in the first
tests for the same field.
With alloy No. 4, Fig. 8, the elongation was about 50 per cent.
greater than that shown in Fig. 5. In both curves there exists a
460
[VOL. XXIV.
J. C. MCLENNAN.
7
:
well marked lag, but in the later measurements, the rod returned to
its original length whereas in the earlier ones it showed an elonga-
tion when the field was removed.
It is evident from these measurements that considerable modifica-
tions must have taken place in the structure of the alloy in the
interval which elapsed between the two tests, and that the lapse of
time is a factor as well as changes in temperature and magnetization
in producing alterations in the molecular groupings in the alloy.
TABLE IV.
Alloy No. 2A₁.
dl
dl
H
• I07
H
107
Į
2
0
3.3
445
2.9
174
3.3
330
2.9
425
3.75
284
2.5
745
4.1
180
1.7
670
3.75
104
0.8
635
3.3
Alloy No. 2A,.
118
0.8
280
1.7
705
2.1
425
1.7
155
0.8
0
0
Alloy No. 3A1.
42
2.5
910
27.5
86
8.3
600
25
138
12.5
405
23.3
180
15
320
21.7
215
16.7
280
20.8
312
20
215
19.2
390
21.7
121
15
460
22.5
75
8.3
545
24.2
42
5.8
705
25
0
4.2
890
26.7
No. 6.]
46I
MAGNETIC PROPERTIES OF HEUSLER'S ALLOYS.
TABLE IV. Continued.
Alloy No. 3A2.
50
2.5
705
20.8
75
4.2
880
22.5
104
6.7
635
20.8
125
8.3
405
19.2
155
10
270
16.7
185
12.5
175
12.5
238
14.2
104
8.3
320
16.7
60
4.2
398
18.3
0
0
495
19.2
Alloy No. 3B₁.
53
1.7
740
23.3
80
4.2
592
22.5
110
5.8
425
20.8
138
8.3
370
20
192
12.5
326
19.2
225
14.2
284
17.5
277
16.7
255
16.7
348
18.3
208
15
372
19.2
174
12.5
458
20.8
118
8.3
565
21.7
68
4.2
610
22.5
0
1.7
Alloy No. 3B,
55
1.7
545
21.7
85
4.2
390
20.8
134
8.3
348
20
187
12.5
327
18.3
275
16.7
255
16.7
390
19.2
187
12.5
480
20.8
125
8.3
665
22.5
68
4.2
760
23.3
0
0
Alloy No. 4A1.
112
3.3
565
7.5
208
4.2
452
7.5
270
5
320
7.5
362
5.8
215
7.5
390
6.7
103
6.7
478
6.7
0
5
705
7.5
462
[VOL. XXIV.
J. C. MCLENNAN.
TABLE IV.-Continued.
Alloy No. 4A2.
50
1.7
362
2.9
110
2.5
255
2.5
362
3.3
118
2.5
552
3.4
70
2.1
610
3.4
42
1.25
590
2.9
0
0
The measurements made with the duplicate set of rods are given
in Table IV., and curves representing them are given in Figs. 9, 10,
II, 12, 13, and 14. For convenience in description the rods used
in this investigation are designated as No. 1, A, No. 2, A, No. 3, A,
and No. 4, A, and where subscript figures are used, for example in
No. 3, A₁, and in No. 3, A, they represent the first and second tests
respectively made with the specimen. Alloy No. 3, B is a rod of
another casting made over a year ago from the same metals and
with the same proportions as the pair No. 3 and No. 3, A.
2
With No. 1, A no elongation was observed which accorded ex-
actly with the behavior of the twin rod No. 1.
With No. 2, A, Fig. 9, the results obtained were very similar to
those obtained with rod No. 2. Curve No. 2, A, represents the
elongations obtained with the first magnetization. As the field was
1

42,107
5
No.2 -A,
No.2 - A₂
100 *200** *300** 400
600
800
700
800 900
H.1
Fig. 9.
gradually increased to 750 units the rod steadily lengthened, and
as the field was decreased the rod shortened, and finally when the
field was removed entirely it remained slightly shorter than its
original length. It was then repeatedly magnetized and demag-
netized and afterwards put through the usual cycle of fields.
curve No. 2, A, represents its behavior after this treatment. The
The
No. 6.] MAGNETIC PROPERTIES OF HEUSLER'S ALLOYS. 463
rod as will be seen had assumed a steady state, and the upward
and downward curves coincided exactly. No hysteresis action or
shortening under high fields was observed.
With alloy No. 3, A the measurements made with the first mag-
netization are given in Fig. 10.
Here it will be seen a lag was

25
20
15
10
5
· 100 200
300 400 500 600 700 800 900
Fig. 10.
obtained in the return portion of the cycle, the rod ending with a
slightly greater length than it had originally.
Curve No. 3, A2, Fig. 11, shows its behavior after being repeat-
edly magnetized. Here again a lag was observed in the return
210?

20
15
JO
5
100 200 300
400 *500 600
700 800
Fig. 11.
portion of the cycle, but the shortening referred to under heading
No. 3 page 455 had entirely disappeared, so that the initial and
final lengths of the rod were the same.
The first and second tests with alloy No. 3, B are shown in Figs.
12 and 13. Here again it will be seen, precisely the same behavior
464
[VOL. XXIV.
J. C. MCLENNAN.
was obtained as with No. 3, A. As with the latter a lag was again
observed in the return portion of the cycle and a final lengthening
of the specimen occurred. In the second tests, Fig 13, made after

47,107
20
15
10
1
'100
200 300 400 500 600
Fig. 12.
repeated magnetizations, this change of length disappeared, and the
rod after being put through its cycle returned to its original length.
With alloy No. 4, A, the behavior was exactly the opposite of
that obtained with No. 2, A. In the initial test, No. 4, A₁, Fig. 14,
the rod steadily lengthened as the field was increased to 700 units
47x107

20
15
10
5
100
200 300
500
400
600
700
Fig. 13.
and afterwards as the field was decreased a lag occurred and the
rod ended somewhat longer than it was at the commencement of
the cycle. After it had been repeatedly magnetized and demag-
netized it was again put through a cycle, and curve No. 4, A₂ was
obtained. Here again the values in the downward portion of the
cycle coincided exactly with those obtained in the forward portion
and no lag and no shortening was observed.
No. 6.] MAGNETIC PROPERTIES OF HEUSLER'S ALLOYS. 465
In the measurements with these rods, the greatest elongation
was obtained with samples No. 3, A and No. 3, B. With them the
maximum elongation observed was about 22.5 ten-millionths which
is somewhat greater than that obtained with the specimen No. 3
where the maximum elongation obtained was 16 ten-millionths.
The maximum elongation obtained with rod No. 4, A was some-
what less than that obtained by Mr. Dawes with rod No. 4. With
the former specimen the maximum elongation was about 3 ten-
millionths while that obtained with the latter was over three times
as great. This discrepancy was very likely due to a difference
in the composition of the -

$
10
two rods which should be 10?
brought out by a chem-
ical analysis of the speci-
mens. It is interesting
to note that with both
sets of rods the maximum
elongation was obtained
when the composition of the
alloy was approximately
the same as that of the
5
No.4-A,
No4-A2
* 100 200 300™ 400 *500 600 H
Fig. 14.
specimens which were found by Heusler to give the highest
permeability.
Summarizing the results of the two measurements it would ap-
pear that the structure of the alloys when freshly cast is very
unstable, and that it undergoes a rapid change under repeated mag-
netizations and a slower change under the lapse of time.
In their experiments on magnetostriction Guthe, and Austin state
that no after-effect either in expansion or in contraction was observed
even with the strongest fields, whereas Austin in his first communi-
cation refers very definitely to the observation of such an effect in
his measurements. The explanation probably is that in the first
investigations the rods were in the same condition as those with
which the initial tests described in this paper were made; while in
the later investigation the rods tested (which were obtained from
Dr. Heusler) had very likely been subjected to such conditions
prior to the test as to bring them into the stable condition.
466
[VOL. XXIV.
J. C. MCLENNAN.
III. PERMEABILITY.
(EXPERIMENTS BY MISS L. B. JOHNSON.)
During the period covered by the experiments on magneto-
striction, a series of measurements was made by Miss L. B. John-
son on the permeability of the four alloys. Two rings had been
made at each casting, in addition to the two rods mentioned above,
and with these the experiments on permeability were carried out.
The magnetization was measured by the ballistic method, by revers-
ing the magnetizing force, and the arrangement of apparatus was
similar to that used by Ewing and Klaasson.¹
TABLE V.

Designation
of Ring.
Height.
Thick-
ness.
Mean
Diam.
Cross Section. Turns in
Primary.
Turns in
Secondary.
No. I.
1.56 cm.
.881 cm.
No. II.
1.47 "
.855 "
No. IIIa.
2.44 "
.623 "
No. IIIb.
2.37 ..
.68
((
6.59 cm.
6.615"
6.98
6.87
1.374 sq.cm.
155
100
1.256"
66
153
100
66
1.516 "
((
150
50
66
1.61
แ
115
50
No. IVa.
2.39 "
.703 "
6.86
1.684 "
""
157
50
No. IVb.
1.447"
.838 ..
6.64
66
1.212 ..
"
136
46
TABLE VI.
Alloy No. I.
H
B
3
H
B
1.39
1.39
1.0
14.40
18.7
2.82
4.01
1.4
28.22
3.93
5.35
1.3
37.63
37.8
50.8
7.52
10.7
1.4
86.45
78.9
m

1.3
1.3
1.3
1.3
TABLE VII.
Alloy No. II.
H
B
H
B
M
.814
1.464
1.7
17.5
27.8
1.5
1.66
2.92
1.7
42.5
68.8
1.6
4.99
8.78
1.7
69.3
106.8
1.5
8.55
13.17
1.5
¹ Phil. Trans., Vol. CLXXXIV., A, p. 987, 1893.
No. 6.] MAGNETIC PROPerties of HEUSLER'S ALLOYS. 467
TABLE VIII.
Alloy No. IIIa.
H
B
μ
H
B
μ
.68
6.442
9.3
17.1
201
11.6
1.8
17.1
9.8
28.3
332.8
11.7
4.4
46.1
10.3
36.9
364
9.8
7
71.2
10.1
42.9
413.8
9.6
11.9
129.8
10.8
59.3
483.1
8.1
The dimensions of the rings which have been designated as I.,
II., IIIa, IIIb, IVa, IVb for the purpose of identifying them with
the rods which were denoted by the same numbers, are given in
Table V. The table also contains the number of turns in the
primary and secondary coils of each ring.
TABLE IX.
Alloy No. IIIb.
H
B
М
H
B
.99
4.596
4.6
11.24
70.17
6.2
1.45
9.138
6.3
23.4
160.8
6.8
2.04
11.42
5.5
30.81
221.5
7.1
3.08
18.27
5.9
50.75
352.1
6.9
6.01
34.26
5.7
!
The first set of measurements was made with the rings, just as
they were cast, and the results are given in Tables VI., VII, VIII.,
IX., X., XI. The permeability curves are given in Fig. 15.
TABLE X.
Alloy No. IVa.
H
B
1
μ
H
B
μ
.2745
13.1
48.1
8.692
565.7
65.1
.5032
30.5
58
16.47
933.9
56.7
1.464
96.1
65.6
34.31
1487.2
43.3
2.379
157.2
66.1
45.29
1724.08
38
4.575
314.5
68.7
69.54
2072.3
29.8
468
[VOL. XXIV.
J. C. MCLENNAN.
TABLE XI.
Alloy No. IVb.
H
B
H
B
M
.0379
3.108
82
16.45
1366.93
83.0
.147
13.67
92.9
27.14
2113.44
77.8
.452
49.72
110
34.55
2257.76
65.4
1.192
136.75
114.7
41.77
2424.24
58
2.467
279.72
113.3
51.00
2548.4
49.8
3.331
360.52
108.2
55.93
2586.4
44.4
5.264
535.81
101.7
80.61
2735.04
33.8
7.81
739.70
94.7
92.13
2735.04
24.6
From these it will be seen that Alloys IVa and IVb showed
the highest permeability. Alloys I. and II. were but feebly magnetic,
and Alloys IIIa and IIIb only slightly so.

105
90
75
60
45
30
15
VI
IV
II
10
20
30
40
50
60
70
H
Fig. 15.
It had been expected that the permeabilities of the rings would
give an indication of the relative permeabilities of the corresponding
rods used in the measurements on magnetostriction. The very
considerable difference in the permeability of the two rings No. IVa
and No. IV, which were made at the same casting, showed, how-
ever, that even different portions of the same casting did not exhibit
the same magnetic properties, and that before any definite com-
parisons could be made, it was necessary to make a chemical
No. 6.]
MAGNETIC PROPERTIES OF HEUSLER'S ALLOYS. 469
analysis of each of the specimens tested. This difference in mag-
netic properties was also very marked with alloy No. III. The
rods of this casting gave the greatest elongation, while the rings,
as the figures show, possessed but a small permeability. Guthe
and Austin¹ have pointed out that the greatest elongation was
obtained by them with rods which possessed the greatest permeability
and as some experiments which are now being made on the permea-
bility of the rods No. III. show that they possess a very much
higher permeability than that exhibited by the rings, this conclusion
is very probably correct.
TABLE XII.
Alloy No. IIIa.
H
B
μ
H
B
३
.73
8.59
11.9
17.1
309.2
17.9
1.28
15.03
11.7
27.5
440.1
15.9
1.97
25.77
13
36.1
541.1
14.9
6.4
94.4
14.7
46.4
674.3
14.5
6.8
105.2
15.3
64.5
751.6
11.6
8.5
138.2
16.2
91.1
843.9
9.2
11.5
196.5
17
Alloy No. IVa.
H
B
μ
H
B
.164
9.83
59.9
8.88
758.02
85.3
.814
65.33
80.5
16.4
1187.42
72.4
1.41
121.02
85.8
34.31
1857.5
54.1
2.33
203.15
87.1
43
2090.1
48.8
4.3
415.05
96.7
71.32
2402.6
33.6
6.03
538.26
89.2
Alloy No. IVb.
H
B
μ
H
B
μ
.7621
10.567
13.9
24.38
528.22
21.6
1.258
28.172
16.3
31.24
654.99
20.9
3.35
56.34
16.4
44.2
795.85
18
6.09
116.2
19
58.68
993.06
16.9
9.9
211.29
21.3
86.11
1253.6
14.55
15.24
352.15
23.1
1 Loc. cit.
470
[VOL. XXIV.
J. C. MCLENNAN.
.
In addition to the initial tests just described, some measurements
were also made on the changes in the permeability produced by
continued heating.

jul
15
10
I
5
90
80
70
60
50
10 20
30
40
50
60
70
80
90 H
Fig. 16.

40
II
30
20
10
10
20
30
40
50
60
70
H
Fig. 17.
Examples of these measurements with alloys No. IIIa and No.
IVa and No. IVb, are given in Table XII.
No. 6.] MAGNETIC PROPERTIES OF HEUSLER'S ALLOYS. 47I
The values of the permeability of alloy No. III. were obtained when
the specimen had become cool after being heated to 107° C. and
kept at this temperature for 40 hours. A permeability curve repre-

ё
105
90
75
60
45
30
15
II
I
10 20
30
40
50 60 70 H
Fig. 18.
senting the results is given in Fig. 16, and a comparison with the
curve made in the initial test which is plotted in the same figure
shows that although the magnetic properties of the specimens were
ё
'

120
110
100
90
80
70
-200
•160
-120
-80-40
O
+40
+80
°C
Fig. 19.
not very marked the heating nevertheless produced a very consid-
erable increase in the permeability.
472
[VOL. XXIV.
J. C. MCLENNAN.
This increase in the permeability is also shown by the numbers
for alloy No. IVa. Curve I., Fig. 17, shows the permeability of the
alloy in different fields when freshly cast, and curve II. its permea-
bility in the same fields after it had been repeatedly heated up to
100° C. and cooled to the temperature of liquid air.
The results obtained with alloy No. IVb show the effect of rais-
ing the alloy to a red heat and then allowing it to cool slowly.
This specimen in the initial tests showed the highest permeability
of all the rings, but after being treated as described a very consid-
erable decrease in the permeability ensued. Curves I. and II. rep-
resent respectively the initial and final tests with this specimen.
In the interval between which the results with alloy IVa recorded
in Fig. 17 were obtained, this specimen was examined under a con-
stant magnetizing force of 3.934 C.G.S. units for temperatures rang-
ing from — 182° C. to 105° C. The temperatures were estimated
from variations in the resistances of a platinum wire wound round
the ring between the primary and secondary coils. The results of
these measurements are given in Table XIII. and a curve repre-
senting them is given in Fig. 19.
TABLE XIII.
Alloy No. IVa. H=3.934 C.G.S. units.
Temperature.
B
Temperature.
B
μ
-182
487
123.7
0
338.5
86
-169
471.8
119.9
8.2
334.2
84.9
-154
454.3
115.4
19.6
327.6
83.2
-143.5
445.6
113.2
35.3
321.1
81.6
-131
432.5
109.9
48
308.8
78.5
-110.5
410.6
104.3
63.1
297
75.5
-53
375.7
95.4
69
288.3
73.2
-46
364.8
92.7
76.4
281.8
71.6
-36
358.2
91
91.6
266.5
67.7
-13.5
342.9
87.1
105
521.21
63.8
The permeability, it will be seen, showed a steady increase as the
temperature was lowered.
The results of these various tests, it will be seen are in accord-
ance with the observations of other investigators, who have found
that the permeability of these alloys can be considerably increased
by continued heating at moderate temperatures.
No. 6.]
MAGNETIC PROPERTIES OF HEUSLERS ALLOYS. 473
The results obtained with alloy No. IVa lends support to the
view expressed by Hill, that when these alloys are raised to a tem-
perature higher than the transformation point, and then cooled, they
possess a permeability which is largely determined by the temper-
ature from which the cooling took place. This alloy when cooled
from the melting point as curve I., Fig. 18, shows, possessed a per-
meability very much higher than that which it had when cooled
from a red heat.
This suggests a parallel between the behavior of these alloys
and those studied by Heycock and Neville. In their investigations
it was found that the structure of an alloy at any temperature could
be ascertained very closely by suddenly chilling it from that tem-
perature. It is possible that these alloys under varying condi-
tions of temperature, may pass through different phases, some of
which may be magnetic, and others not. From the general results
of the present investigations it is evident that the magnetic properties
of the different alloys are intimately associated with their molecular
structure, and since the structure of an alloy at any temperature can
be ascertained by rapid cooling from that temperature, it is possible
that the magnetic properties of the alloy may also be investigated by
the same procedure. The results obtained by Hill and other investi-
gators would seem to point in that direction.
PHYSICAL LABORATORY,
UNIVERSITY OF TORONTO, December 20, 1906.
1 Rapports Congrès Int. de Phys., Paris, 1900, p. 131.
[i
i
1
T
i.
J
!|
ין
!!!
!!
!
!!
li.
1.
!
il
15
1:
1
1
1


UNIVERSITY OF TORONTO STUDIES.
J
The "Papers from the Physical Laboratories," to be issued as a
special series of University of Toronto Studies, date from the year
1900. Nos. 1-17 were published by the Physical Department in
a very limited edition and are no longer in print. For the sake of a
complete record the numbering of the Papers, as forming a series
of University of Toronto Studies, is made continuous with the earlier
series and commences with number 18. The earlier numbers, except
those to which a price is attached, are not now available either for
sale or gift.
No.
No.
No.
1: Electric Screening in Vacuum Tubes, by J. C. Mc-
LENNAN.
Trans. Roy. Soc. of Canada, second series, Vol.
VI, Sec. III (1900).
2: Electrical Conductivity in Gases traversed by Cathode
Rays, by J. C. MCLENNAN.
(a) Phil. Trans., A. Vol. 195 (1900), pp. 49-77.
(b) Zeitschrift f. Physik. Chemie XXXVII, 5 (1901).
3: On a kind of Radioactivity imparted to certain Salts by
Cathode Rays, by J. C. MCLENNAN.
(a). Phys. Zeit. Vol. 2, No. 49, pp. 704-706 1901).
(b) Phil. Mag., Feb. 1902.
No. 4: On excited Radioactivity, by R. M. STEWART.
No.
Trans. Roy. Soc. of Canada, second series, Vol.
VIII, Sec. III (1902).
5: Induced Radioactivity excited in Air at the Foot of
Waterfalls, by J. C. MCLENNAN..
(a) Phys. Zeit., 4, No. 10, pp 295-298 (1903).
(b) Phil. Mag., April 1903.
(c) Physical Review, Vol. XVI, No. 4 (1903).
(d) University of Toronto Studies, Physical Science
Series, No. 1.

0.50
No. 6: Some Experiments on the Electrical Conductivity of
Atmospheric Air, by J. C. MCLENNAN and E. F. BURTON. 0.50
(a) Physical Review, Vol. XVI, No. 3 (1903).
No.
(b) University of Toronto Studies, Physical Science.
Series, No. 2 (1903).
7: On the Radioactivity of Metals Generally, by J. C. Mc-
LENNAN and E. F. BURTON . . . .
(a) Trans. Roy. Soc. of Canada, Second Series Vol.
IX, Sec. III (1903).
(b) Phys. Zeit. 1903, No. 20, pp. 553-556.
(c) Phil. Mag., Sept. 1903.
(d) University of Toronto Studies, Physical Science.
Series, No. 3 (1903).
0.25
No.
8: On the Potential Difference required to produce Elec-
trical Discharges in Gases at Low Pressure; an
Extension of Paschen's Law, by W. R. Carr.
Trans. Roy. Soc. of Canada, Second Series, Vol.
VIII, Sec. III (1902).
No. 9: On the Laws governing Electric Discharges in Gases
at Low Pressures, by W. R. Carr.
Phil. Trans. A. Vol. 201, pp. 403-433 (1903).
No. 10: On the Character of the Radiation from Ordinary
Metals, by E. F. BURTON.
Phys. Rev. Vol. XVIII, No. 3 (1904).
No. 11: The Metric System of Weights and Measures, by
J. C. MCLENNAN.
Proc. of Select Standing Committee on Agriculture
and Colonization of the House of Commons, Canada,
March 29, 1904.
No. 12: On the Radioactivity of Natural Gas, by J. C. MCLENNAN.
Trans. Roy. Soc. of Canada, Second Series, Vol. X,
Section III (1904).
No. 13: On a Radioactive Gas from Crude Petroleum, by
E. F. BURTON.
(a) University Studies, Phys. Sci. Series, No. 4 (1994).
(b) Phil. Mag. Oct., 1904.
No. 14: On the Radioactivity of Mineral Oils and Natural
Gases, by J. C. MCLENNAN.
Proceedings of the International Electrical Congress
of St. Louis, 1904.
No. 15: Note on the Use of Sensitive Quadrant Electrometers,
by J. C. MCLENNAN.
Phys. Rev. Vol. XX, No. 3 (1905).
No. 16: On the Decay of Excited Radioactivity from Natural
Gases, by Miss L. B. JOHNSON.
Phys. Rev. Vol. XX, No. 3 (1905).
No. 17: On the Secondary Radiation excited in Different Metals
by the y Rays from Radium, by H. F. Dawes.
Phys. Rev. Vol. XX, No. 3 (1905).
No. 18: On a New Method of Determining the Specific Heat
of a Gas at Constant Pressure, by H. F. Dawes..
Trans. Roy. Soc. of Canada, Second Series, Vol. XII,
Section III (1906).
No. 19: On the Magnetic Susceptibility of Mixtures of Salt
Solutions, by J. C. MCLENNAN and C. S. Wright
Phys. Rev. Vol. XXIV, No. 3 (1907).
0.25

0.25
+
0.25
No. 20: On the magnetic properties of Heusler's alloys, by
J. C. MCLENNAN
0.25
Phys. Rev., Vol. XXIV.