]\gt ; A Contribution to the Study of the Magnetic Properties of nganese of some Steels .
By Sir ROBERT HADFIELD , F.R.S. , C. , and NEAU .
( Received May 4 , 1917 .
) I.\mdash ; GENERAL AND DETAILED CoNSlDERATION OF THE ATUs AND METHOD MAGNETIC ERTIES .
The object of this research has been to find a connection between the magnetic susceptibility per unit mass ( called the mass susceptibility ) of different anese steels and their chemical constitution .
All measurements were made with the help of the magnetic balance of P. Curie and C. Chenevea In this instrument the body to be tested is attached to one arm of a torsion balance .
A horseshoe magnet can be moved horizontally , so that the direction of motion is at angles to the line joining the poles , and the body is gradually brought into a stronger and stronger field .
The observed deflection of the torsion balance increases at first , then diminishes , and becomes zero when the body lies between the poles of the magnet .
If the magnet be moved still further , a deflection in the opposite direction is observed , which passes through a maximum at the point where the variation of the field of force is greatest .
illustrates the deflections observed in a particular case .
The difference between the readings of the torsion balance when the deflections are greatest on either side is proportional to the mass susceptibility , so that this quantity may be determined by comparison with a specimen whose susceptibility is known .
Throughout the paper the mass susceptibility is denoted by the symbol Glass Envelopes .
Unless otherwise indicated , all specimens were included in glass tubes of very different shapes , and generally with very thin walls .
These are shown in fig. 5 .
They gave an inappreciable deflection in the most intense magnetic fields , even when the balance was at its greatest sensibility , i.e. when the platinum wire was the thinnest procurable .
The of this glass , the search for which has entailed great trouble , is exceedingly advantageous , in that it does away with the necessity of correcting for the magnetic effect of the envelope . .
de Phys Series 4 , vol. 2 , p. 796 ( 1903 ) ; 'Phil .
Mag vol. 20 , p. ( 1910 ) .
Sir R. Hadfield , C. Cheneveau , and Ch. Geneau .
FIG. l.\mdash ; Variation of deflection shown as function of displacement of magnet .
Standard of P. Weiss and have devoted great attention to the determination of the mass susceptibility of salts in solution and in the solid state .
We have adopted , as standard of comparison , anhydrous cobalt sulphate , to which these workers assign the value at C. We have prepared some of this salt , adhering strictly to their instructions .
It was divided into three portions , of which one was placed in a tube , which was then sealed and weighed .
This served as a standard .
The 'Journ .
de Phys Series 5 , vol. 1 , p. 274 ( 1911 ) .
Magnetic Properties of and IIIanganese Steels .
67 following Table shows that this sample , placed in fields of different strengths and under different conditions of sensibility , always gave concordant results .
Sensibility , calculated 180 36 open t Ratio of sensibilities Rntio of deflections Ratio of sensibilities Rntio of deflections Ratio of sensibilities Rntio of deflections Ratio of sensibilities Rntio of deflections Ratio of sensibilities Rntio of deflections Ratio of sensibilities Rntio of deflections Ratio of sensibilities Rntio of deflections Ratio of sensibilities Rntio of deflections Ratio of sensibilities Rntio of deflections Ratio of sensibilities Rntio of deflections Ratio of sensibilities Rntio of deflections Ratio of sensibilities Rntio of deflections Ratio of sensibilities Rntio of deflections Ratio of sensibilities Rntio of deflections Ratio of sensibilities Rntio of deflections Ratio of sensibilities Rntio of deflections Ratio of sensibilities Rntio of deflections 4.8 Two permanent magnets ( Allevard steel ) , of the following dimensions , were used to establish the field of the Curie balance .
The ratio of the fields was about 5 .
gnet I. gnet Il.\mdash ; Exterior diameter Interior Thickness Thickness of pole pieces Height of polar space Breadth of The values of the field strengths , which are only approximate , were calculated from the general formula of the apparatus by the aid of the known value of for cobalt sulphate and from the constants of the torsion balance .
An excellent verification of the validity of this procedure was obtained in Case II by the aid of a Grassot fluxmeter .
This maxwells .
Taking an approxitnate value of 12 for the permeability of steel , and sq .
cm .
for section of polar space , a field of 1750 gauss was cal culated .
To obtain more feeble fields the magnets were shunted , but the results of such tests were very irregular .
Sir R. Hadfield , C. Cheneveau , and Ch. Geneau .
Sensibility .
Varying sensibilities were obtained not only by interchanging the magnets , but also by altering the diameters of the platinum suspending wires .
The .
relation of the sensibility to the diameters is given in the following Table:\mdash ; Wires. .
Sensibility ith\mdash ; Magnet , , II . .
, .
Wire No. 1 broke very early on in the work , and was replaced by No. 4 .
No. 3 gave very irregular results .
The most trustworthy results have been obtained with Nos. 2 and 4 .
The effect of the susceptibility of the air was found to be igible in all cases .
Influence of Tpatnre.\mdash ; The following experiments made on steel when placed in a glass tube showed that in these experiments a variation of C. had but little effect on the results .
Steel No. 1010 N. Mass , .
Distance of scale , 236 cm .
Accuracy of In all cases , on account of experimental difficulties , and the many sources of error which had to be overcome , the steel results can only be trusted to within 5 per cent. The manganese results are a little more accurate .
Their precision is indicated in the results .
Magnetic Properties of nganese Manganese Steels .
69 \mdash ; STUDY OF THE MAGNETIC PROPERTIES MANGANESE .
Through the kindness of M. P. Weiss , we have been able to work on the specimens of manganese used by M. Onnes and himself in their work on the netic properties of manganese , chromium , and vanadium .
* ( 1 ) Powdered The first speciluen of anese was in the folm of a grey powder .
It was prepal.ed from an amalgam of anese , which , when heated to a sufficient temperature in pure , yielded the metal in the form of the powder mentioned above .
This powder was pressed into a ylass tube in order to form a rod , and then sealed .
The two previous have shown that this form of anese i , s netic .
After a ative confirmation of this view we proceeded to a quantitative test by determining the deviation of the sealed tube in fields I and II .
Unfortunately , we were to break this tube ( naturally with reat care ) in ordel to determine the mass of the anese and the magnetic quality of the glass .
The following results were obtained , a very small glass correction having been allowed for .
gives the number at C. Our number agrees with that found by H. du Bois and Honda , while Gebhardt 's value , quoted by Weiss and Kamerlingh is five times too large .
Without doubt their samplss were oxidised .
We have repeated the experiment upon the manganese powder after it had been kept for 10 days in a closed tube .
We have obtained There was therefore no sensible oxidation .
* Pierre Weiss and Kamerhngh Onnes .
' ' Comptes Rend .
de l'Acad .
des Sci vol. 150 , p. 687 ( 1910 ) .
'Ann .
Physik , ' vol. 37 , p. 657 ( 1912 ) .
Journ. de Phys Series IV , vol. 9 , p. 656 ( 1910 ) .
Sir R Hadfield , C. Cheneveau , and Ch. Geneau .
have assured ourselves by the sensitive potassium thiocyanate reaction that there was no in the sample .
We conclude , then , that the mass susceptibility of powdered anese at C. is within 2 per ceut .
( 2 ) Cast The second sample of manganese was composed of small black pieces , in parts coloured a metallic or yellow .
It was prepared by the fusion of the powder in an electric resistance i.urnace , the atmosphere being drogen .
Weiss and Onnes state that this form is ferro-magnetic .
these pieces we carried out a series of experiments which showed that the substance was either attracted or repelled , according to its position in the field .
This obSeryation , which was verified for each separate piece , indicates that this sample of was ynetised .
A , but sensible , effect upon a compass needle confirmed this .
Five years and five months after our first experiments we again tested this cast manganese , and found that it still retained its magnetism\mdash ; which shows a strong coercive force .
We then carefully heated two yments to a bright red\mdash ; one a fra meant of the black ( No. 1 ) , the other a fragment with metallic lustre ( No. 2 ) .
The following results were obtained .
Initially the hest values of were of the order Black meant Brilliant fragment The influences of heating , of temperature , and of the size of the sample , are of some importance .
With a fairly large fragment ( No. 3 ) we obtained M. Kuh Zurich , has given for the magnetic moment of cast anese : \mdash ; lt is to be observed that for a field strength of about 320 gauss we have found numbers of the same ordel for the highest values of of ferro-magnetic manganese , viz. , 2200 and , after an interval of five years , 2700 .
netic Properties of in pure rave successively 1st 1st 2nd , , .
and .
in 2nd , 1st is therefore clear that sufficiently prolonged lessens the uetic properties of magnetic cast anese , whilst the oxidation the specimen slightly raises its susceptibility .
In .
specimen No. 1 above in ordel to it , without in acid , we obtained while with specimen No. 2 , after in acid , we obtained suc , cessively Initial } 1st washing The effect of and vashing is clearly to cause an approach to the value for ered manganese .
) tion ; snbsequent washing in acid the disappearance of the oxide forlued .
Desirous of the remainder of the nese , further erintell were discontinued .
If the anese dnetised by and deoxidised by the action of hydrochlori acid were ferro-magnetic , it should easily -acquire its netism when placed in the field of No. 1 , or even by simply into contact with another magnet .
this is not so , is shown by the results , giving identical values before and after the attempted metisatiou .
high coercive force of manganese fused in ( analogous to that of iron under similar conditions ) and the preceding expel.iments lead us to the belief that its ferro-magnetism is due to hydrogen .
The difhculty of Sir R Hadfield , C. Che'neveau , and Ch. Ge'neau .
completely a dissolved gas from a liquid is well known , and the same difficulty must arise with a solid .
It is hardly to be expected that the heat applied to our samples was sufficient to drive off all Ghe hydrogen , and so to lower the msgnetic properties of the to that of the pure metal .
But the gas was driven off sufficiently to prevent the metal being permanently magnetised .
This conclusion is verified by an experiment of Seckelson , has obtained electrolytically ferro .netic manganese , which necessarily contains hydrogen just as did the electrolytic iron of Cailletet and Houllevigue .
anese is para-magnetic , whatever its condition .
Its mass susceptibility at C. .
The -magnetic properties which it possesses after being melted or deposited by electrolysis are without doubt due to occluded hydrogen .
It is perhaps in place to point out that the presence of occluded ases in metals or alloys which have been prepared by should not be neglected when ex mining the netic properties .
III.\mdash ; STUDY 0F THE MAGNETIC OPERTIES OF SOME GANESE STEELS .
The first experiments made with steel ( in the form of small cylindrical rods ) gave very irregular results .
For instance , they showed the steel 2084 A as ferro-magnetic , which was subsequently confirmed by experiments , But also gave the steel 1010 as ferro-magnetic , whereas it is in fact para-magnetic .
showed , however , in all cases , in spite of the irregularities , the cause of which we shall try to find , that all the steels save the one mentioned above were only slightly magnetic , and the employment of the type of apparatus used was justified in the overwhelming , majority of cases .
In what follows we confine ourselves to para-magnetic steels , later we will deal with the ferro-n ) agnetic va1iety .
( 1 ) Influence of eagents .
The paradoxical result obtained with the steel 1010 may easily be explained on the hypothesis that some ferriferous impurity gives to the specimen its onetic character .
In fact , if the cylinder is gently washed with hydrochloric acid , the mass susceptibility passes rapidly from a 'Wied .
Ann vol. 67 , p. 37 ( 1899 ) .
' Journ. de Phys Series 3 , vol. 6 , p. 245 ( 1897 ) .
Magnetic Properties of anganese and Manganese .
73 value of too large to be measured to following experiment lends weight to this assumption .
When the cylinder was cut by an iron tool without subsequently washing in acid , rose in value from to .
As the steel bars were worked into shape , and as it is often necessary to test small pieces cut from them , it is therefore essential , before using them in the balance , to wash them in hydrochloric acid , in order to remove traces of iron left by the tools .
An examination of the curves in fig. 2 shows that for the stee11010 , whatever the form of the specimen , the deflectiuns which , before chemical treatment , were not proportional to the weight became so when tlJe sample was so treated .
* Steel lOlOG Sensibility .
Dimensions .
Before attack by IICI .
After .
attack .
tion Time of Diameter .
per .
per gramme Sphere Finally we mention here the observations of A. C. Jolly and of one of upon copper .
In this case the mass susceptibility only became tant after a first washing with hydrochloric aid .
Subsequent washings had little effect .
This is also shown in the curve for the steel 1010 ( iig .
2 ) , although the mass of the specimen was reduced in the process from .
to Alth u it would seem that in many instances the ferriferous impurity resides at the surface , yet there are others in which this does not seem to be entirely the case .
We find it difficult to decide , however , wheCher we are really dealing purely with superficial effect , with an internal effect , or with a combination of the two .
We simply estimate that the diminution in weight due to washing is at its maximum about 1 or 2 per cent. Other experiments upon the steels 1010 , 1010 , 1010 , 1010 1339 , 1343 , 1343 , 598 and 620 , and the curves for the specimens 1010 * The hydrochloric acid used was strictly diamagnetic with .
If true value ( from Quincke 's work ) is , then the acid above does not contain more than 1/ 100,000 per cent. of iron .
' The Electrician , ' vol. , No. 25 , p. 1013 ( 1910 ) .
C. Cheneveau , Journ. de Phys Series 4 , vol. 9 , 163 ( 1910 ) .
Sir R. Hadfield , C. Cheneveau , and Ch. Geneau .
and ( fig. 2 ) Iead equally to the conclusion that , after a fairly prolonged in ochloric acid , the deflections become practically proportional mass of the specimen .
FIG. 2.\mdash ; Influence of acid on sensibility .
( 2 ) of .
of For Jletic bodies in which the susceptibility is not a function of the lield strength , the shape of the specimen should be immaterial and , in consequence , its dimensions should not affect the measurements , otherwise than the of mass .
Certain of the steels were very hard to work and in the absence of diamond dust we shaped them with ordinary tools .
But for our experiments we chose Magnetic Properties of Manganese Steels .
75 fairly workable steels which were somewhat easily attacked by the acid .
The washing was continued until strict proportionality was obtained between deflection and weight .
With the steel 1798 we find that when the sample is properly placed between the pole pieces the deflection is the same whether the specimen is placed parallel to the line joining the poles or perpendicular to it .
With the same steel we carried out a selies of experiments under different conditions of sensibility , in which we varied one of the dimensions of the cylindel\ldquo ; keepin the other constant .
The curves in fig. 3 show the relationship between the deflection and the height and diameter of the specimen .
FIG. 3.\mdash ; Variation of deflection with dimensions .
Experiments were also carried out on a very different type of steel ( 1010.N ) .
The curve ( fig. 3 ) shows , in agreement with our previous results , that , provided certain limits are not exceeded , the deflection.is proportional to the dimensions of the specimen , i.e. , to its weight .
The conclusion to be drawn from these preceding series of experiments is that serious error can only be avoided by working on specimens giving deflections proportional to their masses .
It is , therefore , advantageous to work with coml ) aratively small samples .
VOL. ( c ) of devition with the position of the magnet.\mdash ; This is shown in fig. 1 .
It is easy to deduce from these experiments the position for which small displacements least affect the of the results .
Magnetic Properties of nganese and Steels .
77 ( 4 ) Best Bxperimenta It has been shown that it is advantageous to use fairly small specimens having a well defined geometrical shape ( cylindrical ) .
The specimen must be sufficiently washed in hydrochloric acid , and accurately placed with its centre at the centre of the polar interstice\mdash ; that is , at equal distances from the pole pieces , and vertically , so that the centre of the specimen lies in the central line joining the pole pieces .
This adjustment , which is somewhat delicate , can , nevertheless , be easily accomplished by the employment of tubes having the forms shown in fig. 5 .
The specimen is retained in position by means of ghtly packed cotton-wool .
FIG. 5.\mdash ; Various experimental tubes .
Sir R. Hadfield , C. Cheneveau , and Ch. Geneau .
The following numbers for the specimen show that error caused by small variations of is slight , if the of several experiments be taken .
( 5 ) The of Ferro-magnetic St The magnetic torsion balance only lends itself to the determination of the mass susceptibility of diamagnetic and paramagnetic bodies .
For the successful study of ferro-magnstic bodies , it is essential that the rod should be placed parallel to the field , and that its length should be great in comparison with its diameter .
* We have only been able to work on very small specimens , weighing a few milligrammes , and , under these circumstances , it was impossible to eliminate the demagnetisin field due to the magnetisation .
In all cases , however , we have endeayoured to correct this error by approximate mathematical computation .
Such a correction is naturally somewhat uncertain , and we tested and verified its approximate validity by comparing the specimen 2084 A with a sample of pure iron wire ( piano wire ) .
Washing in hydrochloric acid was , as usual , carried out before testing .
To prevent the tube coming into contact with the magnet of the instrument when it contained ferro-magnetic matter , we attached a weight from a hook fixed to the lower end of the tube , as shown in fig. 4 .
We ascertained that this did not affect the deflections when the weight did not exceed 10 Weight .
Deflection at 164 cm 172 mm. 178 mm. 178 mm. 180 mm. The following are the results ( after all corrections had been applied ) obtained in two perpendicular directions , with the sample placed either vertically and at right angles , or horizontally and parallel , to the field .
The different values obtained in these cases is typical of the ferro-magnetic state .
e.g. , P. Curie , ' ( Euvres , ' 1908 , p. 289 .
ILagnetic operties of Steels .
79 Thus , the stee12084 A is htly ferro-magnetic compared with iron .
Its mass susceptibility is of the same order as that of pure iron .
A curious point arises in comlection with this specimen .
Before being sent to Paris in 1911 , it was roughly examined with a hand magnet , and showed only magnetic quality , certainly not more than 4 or 5 per cent. of that of pure iron .
The above test , however , which was made about six years later , indicates distinct ferro-magnetic properties of the of 50 per cent. of that of pure iron .
This marked difference was so striking that it was well to make a further examination of the specimen in the original approximate manner , with the result that this considerable increase of magnetic susceptibility with progress of time was fully confirmed .
The specimen had not , during an interval of several years , been subjected to heat or mechanical treatment , and it can only be inferred that a gradual secular and internal change in magnetic character , apart altogether from any extraneous influences , had taken place in this material .
( 6 ) Study of Powdered Steels .
These experiments were carried out on filings from samples 1010 and 1109 D. The chemical composition ( which is very similar to that of the rods ) is given below , together with the results of the magnetic tests .
These filings are so magnetic that we found it necessary to use a stouter torsion wire ( diameter mm and to fix a netic shunt to magnet No. 2 .
In a field of about 18 gauss ( 1/ 18 of its previous value ) we obtained the following results:\mdash ; Steel NDefiection pramme * The magnet used here was slightly stronger than that in the other case .
For the same torsion wire , we deduced from deflections made with cobalt sulphate that the field in the position of greatest deflection was 460 gauss .
In a field of 460 gauss it can be shown from Curie 's experiments that Sir R. Hadfield , C. Cheneveau , and Ch. Geneau .
These deflections are approximately 130 times those given by the same material in bars .
Due precautions were taken to place the sample at the centre of the tube ( fig. 6 ) , and we also tested the strict proportionality between weight and deflection ( fig. 6 ) up to the limits to which the experiments were confined .
It is certain that these contain a considerable quantity of iron .
Washing in acid was resorted to , but we never reached a final value nor got rid of all the iron .
With steel 1010 we even found an increase of ation ( , ergesting that the constituents of the powder were attacked by the acid .
We conclude that filings cannot be used for the determination of the mass susceptibility of manganese steels .
IV.\mdash ; SUMMARY OF EXPERIMENTAL RFSULTS .
In the following Tables the chemical and physical results for the special manganese steels are collected together so as to determine the relations between chemical composition and the magnetic properties .
Properties of and Manganese Steels .
81 ( 1 ) of fanganese .
Within the range examined , it appears that in the case of manganese steels of constant carbon content this metal has only a feeble effect on the magnetic properties .
( 2 ) of Carbon .
The Table also shows that for steels of constant manganese content the susceptibility increases as the carbon content diminishes .
These two conclusions are summarised in the column giving as a function of the quotient , and also in the curve representing the figure which is reproduced in fig. 7 .
The mass susceptibility of manganese steel is seen to become smaller as the ratio increases .
Recapitulative Table , showing lnfluence of anese and Carbon .
/ 8@ Sir R. Hadfield , C. Cheneveau , and Ch. Magnetic Properties of Steels .
83 The results obtained in the two following series of experiments are in good agreement with the conclusions arrived at .
In the first the manganese and carbon contents increase approximately in the same ratio:\mdash ; It is seen that remains approximarely the same .
The second series contains steels of constant carbon content and increasing manganese content .
As the increment of manganese is only per cent. , and as we have seen that an increment of per cent. does not affect , it follows that these steels should have the same netic quality .
This is confirmed by the results:\mdash ; ( 3 ) Influence of iVickel .
If the nickel-manganese steels , containing approximately the same carbonmanganese ratio , are arranged in order of their nickel content , it will be seen either from the following Table , or from the curve of fig. 8 , that the mass susceptibility of such steels increases with the proportion of nickel:\mdash ; Sir R. Hadfield , C. Cheneveau , and Ch. Geneau .
( 4 ) of Copp : The addition of copper to a manganese-nickel steel appears to raise the mass susceptibility , as shown below:\mdash ; ( 5 ) Influe of Chromium .
Chromium also seems to raise slightly the mass susceptibility:\mdash ; ( 6 ) Influence of Tungsten .
Tungsten seems to act more strongly , for the addition of per cent. has a ularked effect on the susceptibility:\mdash ; qnetic Properties of IVIanganese and nganese Steels .
85 ( 7 ) of Silicon .
The addition of a considerable proportion ( 6 per cent. ) of silicon to a manganese steel makes it ferro-magnetic when the temperature of is high : We have already seen that this steel resembles iron in its magnetic properties .
V.\mdash ; RELATIONS BETWEEN MAGNETIC PROPERTIES AND OTHER PHYSICAL PROPERTIES .
( 1 ) As was to be expected , there is no simple relation between density and mass susceptibility , just as there is none between density and chemical composition .
( 2 ) Magn jtic Field .
For certain of the steels , especially those of high susceptibility , there are small variations between the values in fields of different strengths .
In , the highest susceptibility corresponds to the weakest field .
In view of the big difference between the fields and the smallness of the variation , it would not be safe to conclude that this is due to feeble netism .
On the contrary , it can be stated that , within the errors of experiment , the mass susceptibility is in the majority of cases independent of the field strength .
The curves of fig. 8 , referring to manganese-nickel steels , in fields 1 and 2 , prove this .
Apart from the manganese-silicon steel , which is decidedly ferro-magnetic , the other steels are para-magnetic .
( 3 ) Heat Treatment .
To show the effect of the temperature at which the steel was cooled , we give a comparison of similar steels:\mdash ; 86 Sir E. Hadfield , C. Cheneveau , and Ch. Geneau .
The nickel-content in these steels is approximately constant , and , as we have seen that independently of the temperature of quenching the auscepti- bility should increase as the value diminishes , it appears from the experiments that the susceptibility is increased by a low temperature of As chromium has ] a feeble effect on the magnetic properties of manganese steels , ibs presence does not explain the difference found below between steels 2103 and 1109 D. \mdash ; teelNo .
Thermal treatment .
86 Sir E. Hadfield , C. Cheneveau , and Ch. Geneau .
The nickel-content in these steels is approximately constant , and , as we have seen that independently of the temperature of quenching the auscepti- bility should increase as the value diminishes , it appears from the experiments that the susceptibility is increased by a low temperature of As chromium has ] a feeble effect on the magIJetic properties of anese steels , ibs presence does not explain the difference found below ween steels 2103 and 1109 D. \mdash ; teelNo .
Thermal treatment .
86 Sir E. Hadfield , C. Cheneveau , and Ch. Geneau .
The nickel-content in these steels is approximately constant , and , as we have seen that independently of the temperature of quenching the auscepti- bility should increase as the value diminishes , it appears from the experiments that the susceptibility is increased by a low temperature of As chromium has ] a feeble effect on the magIJetic properties of anese steels , ibs presence does not explain the difference found below ween steels 2103 and 1109 D. \mdash ; teelNo .
Thermal treatment .
86 Sir E. Hadfield , C. Cheneveau , and Ch. Geneau .
The nickel-content in these steels is approximately constant , and , as we have seen that independently of the temperature of quenching the auscepti- bility should increase as the value diminishes , it appears from the experiments that the susceptibility is increased by a low temperature of As chromium has ] a feeble effect on the magIJetic properties of anese steels , ibs presence does not explain the difference found below ween steels 2103 and 1109 D. \mdash ; teelNo .
Thermal treatment .
86 Sir E. Hadfield , C. Cheneveau , and Ch. Geneau .
The nickel-content in these steels is approximately constant , and , as we have seen that independently of the temperature of quenching the auscepti- bility should increase as the value diminishes , it appears from the experiments that the susceptibility is increased by a low temperature of As chromium has ] a feeble effect on the magIJetic properties of anese steels , ibs presence does not explain the difference found below ween steels 2103 and 1109 D. \mdash ; teelNo .
Thermal treatment .
86 Sir E. Hadfield , C. Cheneveau , and Ch. Geneau .
The nickel-content in these steels is approximately constant , and , as we have seen that independently of the temperature of quenching the auscepti- bility should increase as the value diminishes , it appears from the experiments that the susceptibility is increased by a low temperature of As chromium has ] a feeble effect on the magIJetic properties of anese steels , ibs presence does not explain the difference found below ween steels 2103 and 1109 D. \mdash ; teelNo .
Thermal treatment .
The conclusion to be drawn is the same as before .
The lower the temperature from which the steel is cooled , the more are the magnetic qualities accentuated .
\mdash ; GLNERAL CONCLUSIONS .
We have determined , with the ynetic balance of P. Curie and C. Cheneveau , the mass susceptibility of manganese and of some of its special netic alloys .
1 .
We have studied manganese under two forms , powdered and fused , in an atmosphere of hydrogen .
The specimens were very kindly lent us by M. P. Weiss , who had prepared and examined them .
* We belve ourselves to bejustified in concluding that whatever its condition , so long as there is no occluded gas , manganese is para-magnetic ; its mass susceptibility at C. is per cent. This value is a little higher tha1r that given by H. du Bois and Honda .
With regard to the netism of cast manganese , we attribute it to the presence of hydrogen .
This agrees with the observation that this type of manganese , after beiIlg heated red , cannot be remagnetised , and with the well known analogous properties of electrolytic manganese and electrolytic ilon .
In this c.onnection we desire to recall the effect which certain occluded gases have upon the magnetic properties of cast metals and alloys .
Weiss and Kamerlingh Onnes , ' Comptes Rendus , ' vol. 150 , p. 687 ( 1910 ) .
Magnetic Properbles of and Manganese Steels .
87 2 .
Special Manganese We have shown that the mass susceptibilities of these alloys , which vary between and ) , are independent of the magnetic field .
The strength of the field from 320 gauss to about five times that value .
These alloys are therefore netic .
( 1 ) ga Steels \mdash ; Within the range examined , the mass susceptibilities are only slightly influenced by the amount of anese present , but depend considerably on the carbon c , ontent .
enerally , we say that for these special steels the mass susceptibility increases as the ratio diminishes .
( 2 ) Steels containing oth Jfetals\mdash ; lf the carbon-luanganese ratio is kept constant , the addition of other metals always increases the mass susceptibility .
Ma tickel Steds.\mdash ; An increase of the nickel ( from to 19 per cent. ) raiscs the mass susceptibility in the atio of 23 : Mangancsc-tungsten Steets.\mdash ; An increment of per cent. of tungsten increased the mass susceptibility in the ratio of 18 : 29 .
Steels.\mdash ; Chromium seems also to cause a increment .
An increase of per cent. of chromium seems to correspond to an approximate increase in the mass susceptibility of 10 per cent. -copper Steel.\mdash ; The comparison of two steels each having the same vanese , nickel and carbon content , one containing no copper , and the other per cent. , shows that ) metal , in spite of.its diamag , increases the mass susceptibility by about 19 per cent. All the preceding steels were quenched in water from 1050o C. They are all of the well known para-magnetic Austenitic type of steel .
The addition of a substance like manganese , which is also para-magnetic , will have but little effect in altering the properties of these steels , which remain paramagnetic .
-silicon Steel.\mdash ; The addition of 6 per cent. of silicon to a manganese steel , where the ratio is about , makes the steel much more ferro-magnetic .
A secular change in the magnetic properties of this steel has also been observed , the specific magnetism increasing , in a period of several years , from the order of 4 or 5 per cent. of that of iron up to nearly 50 per cent.

