1.1. From the Introduction.

The magnetic properties of iron, steel, nickel, and cobalt at moderate and high temperatures have been investigated by numerous experimenters, perhaps the most notable work being that carried out by Hopkinson, M Curie, and D K Morris. Hopkinson employed the ring method of Rowland, insulating the windings from each other by means of asbestos, and deducing the temperature from the electrical resistance of a platinum wire wound upon the specimen. He obtained magnetisation curves, at temperatures lying between ordinary room temperature and the critical temperature of the material, for soft iron, mild steel, hard steel, nickel, and cobalt. In the case of soft iron he found that for low values of the magnetising force the effect of increasing the temperature was to bring about an increase in the permeability; as the temperature approached the critical temperature of the material an enormous increase took place in the magnetic quality. For a value of the magnetising force of 0.3 c.g.s. units the permeability was about 400 at room temperature; as the temperature increased the permeability steadily increased, and at 600° C. had attained a value of 900. From this point on, the increase in permeability with temperature became more rapid; at 700° C. the permeability was about 1900, at 750° C. it had the value 4000, and at 775° C. it reached the maximum value of 11,000. Further heating brought about a very rapid loss of magnetic quality, and at the temperature of 790° C. the iron had become practically non-magnetic. At this temperature its permeability was about unity. For large fields the permeability remained practically constant until a temperature of 600° C. was reached ; there was then a steady falling off in magnetic quality up to 790° C., at which temperature the iron became practically non-magnetic.

Specimens of mild and hard steels exhibited much the same general behaviour; the critical temperature was found to be 740° C. for the former and 690° C. for the latter material.

1.2. Summary.

1.2.1. Experiments of Hopkinson, Curie, and D K Morris.

1.2.2. Precautions necessary in magnetic testing. If the temperature of a test-specimen is changed, the specimen must be rendered neutral at the new temperature previous to carrying out the test.

1.2.3. The effect of increasing the temperature of a specimen of cast iron from 15° C. to 190° C. is to improve its magnetic quality for moderate values of the magnetising force; from 190° C. to 260° C. the magnetic quality falls off, after which there is a further improvement. Finally the specimen becomes non-magnetic at the critical temperature. These results indicate the presence of a transformation point for carbon alloys of iron in the neighbourhood of 200° C.

1.2.4. Steel containing 1·64 per cent, of carbon resembles cast iron in its behaviour. The set-back in magnetic quality takes place at much the same temperature as that which occurs in cast iron, the magnitude of the set-back being much the same for both.

1.2.5. Steel containing 0·8 per cent, of carbon exhibits the transformation point, but in a less marked degree. The first maximum point in the susceptibility curve occurs at 180° C, and the first minimum point at 220° C.

1.2.6. Low carbon steel containing 0·3 per cent, of carbon shows the transformation point; it is, however, very much less marked than in the case of the steels containing the higher percentages of carbon.

1.2.7. The susceptibility-temperature curve of a specimen of soft iron containing 0·06 per cent, of carbon exhibits no turning points in the neighbourhood of 200° C. for fields lying between 2 and 15 c.g.s. units.

1.3. Affiliation.

Carnegie Research Scholar in the University of Glasgow.
Communicated by Professor A Gray, F.R.S.

1.4. Acknowledgement.

The work described in the present paper was carried out in the
Natural Philosophy Institute of the University of Glasgow. The author
desires to express her thanks to Professor Gray and to Dr J G Gray
for the interest they have taken in the progress of the experiments.

2.1. Introduction to the paper in Moir's D.Sc. thesis.

In discussing the method of procedure employed in obtaining I-H curves at various temperatures for a series of carbon steels, I have mentioned the peculiar state of sensitiveness to magnetism which is brought about in a specimen when it is heated. So far, the peculiar condition has been regarded simply as a source of error, and the only concern with it has been to see to its complete removal. Circumstances already mentioned, however pointed to the desirability of a careful investigation of the phenomenon, and such an investigation I accordingly decided to undertake.

2.2. Extract from the paper.

It has been pointed out by several experimenters, notably by Ewing, and Gray and Ross, that a specimen of steel freshly annealed is in a peculiar magnetic condition. Thus, if a specimen, thoroughly demagnetised, is annealed and then tested, a certain I–H curve is obtained. If it be then demagnetised, and tested again, a second I–H curve is obtained which lies definitely below the first; and any further tests after demagnetisation produce a repetition of the second curve, which is characteristic of the specimen. To obtain the first curve again, the specimen has to be annealed from the same temperature as before, and a test made before demagnetisation.

3.1. From the Introduction.

The influence of temperature upon the magnetic properties of iron, steel, nickel, and cobalt has been investigated by numerous experimenters, perhaps the most notable work being that carried out by Rowland, Baur, and Hopkinson.

It is now well known that the effect of raising the temperature of a specimen of iron or steel is to increase its susceptibility to magnetism for low values of the magnetising force and to diminish it for high values. It is to be expected then that, if the temperature of a test-piece be lowered, the effect should be exactly reversed, and the susceptibility should be diminished for low and augmented for high fields. That such is actually the case has been shown for various magnetic materials by Honda and Shimizu, and more recently by Gray and Ross, who have all used the temperature of liquid air, -190° C, as the low temperature at which they have made their investigations. They have shown also that, while alike in this one respect, different materials differ very considerably both in the amount of change in susceptibility brought about by the change of temperature, and in the value of the field-strength for which the effect reverses its sign.

In any such examination, particular interest always attaches to the observation of the effect of similar treatment on the different members of a graded series, either of steels or alloys, the changes in effect being directly due to the changes in the content of the specimens under examination.

I accordingly decided to examine magnetically a specially prepared series of cobalt-manganese alloys, containing respectively 5, 10, 15, 20, 25, and 30 per cent, manganese, the rest of the content of the specimen being pure cobalt.

3.2. Affiliation.

Carnegie Research Fellow in the University of Glasgow.
Communicated by Professor A Gray.

3.3. Acknowledgement.

In conclusion, I desire to express my thanks to Professor Gray and to Dr J G Gray for the interest they have taken in the progress of the work.

4.1. From the Introduction.

The magnetism of permanent magnets is a subject of wide interest and of considerable importance, the value of many different machines depending first and foremost on the constancy of permanent magnetism.

It is a well-known fact that of the residual magnetism which a specimen holds when the applied field is reduced to zero, only a part is held with such a degree of firmness that it can be considered permanent. In the case of some steels, the application of a very small negative magnetising force is sufficient to reduce the residual magnetism to zero, while in the case of others a very considerable negative field must be applied. This negative field is equal and opposite to what is known as the coercive force, which can therefore be considered as a measure of the firmness with which the residual magnetism is held. The hysteresis curve of a specimen then, in indicating the values both of the residual magnetism and the coercive force, gives us a considerable amount of information regarding its permanent magnetism.

Professor Silvanus P Thompson, in a paper read before the Institution of Electrical Engineers, has collected and described the results obtained by various workers on the subject. He shows first of all that, while in the preparation of a satisfactory magnet the material of which it is to be composed is naturally of primary importance, the shape and size must also be considered, and in addition the heat treatment to which it is subjected before magnetisation, a steel hardened by rapid quenching from a high temperature having in general a much greater coercive force than a specimen of the same steel cooled more slowly.

4.2. Affiliation.

Carnegie Research Fellow in the University of Glasgow.
Communicated by Professor A Gray.

4.3. Acknowledgement.

The work described above was carried out in the Natural Philosophy Institute of Glasgow University, and my thanks are due to Professor Gray and to Dr J G Gray for the interest they have taken in the progress of the research, and for valuable advice and suggestions.