]\gt ; [ 247 ] VIII .
The Influence of olecutar Constitution and Temperature on Iagnetic Susceptibility .
Part IV .
\mdash ; Further Applications of the Molecular Field .
By Major A. E. OXLEY , R.A.F. , M.A. , D.Sc .
, Mackinnon Student of the Royal Society .
Communicated by Prof. J. W. NICHOLSON , F.R. Received May \mdash ; Read June 26 , 1919 .
CONTENTS .
Page ( 1 ) Introduction . . . . . . . . .
247 ( 2 ) On the energy and ultimate tensile strength associated with crvstalline media or gels 250 ( 3 ) A comparison of the elastic constants of diamagnetic , paramagnetic , and ferro-magnetic crystalline media . . .
, . . . . .
257 ( 4 ) On the change of volume on crystallization interpreted as a magneto-striction effect of the molecular field . . . . .
259 ( 6 ) Further discussion of the nature of the moleculal field . . . . . .
267 ( 6 ) On a magnetic theory of chemical combination . . . . .
277 ( 7 ) On some anomalies of the Faraday effect . . . . .
282 ( 8 ) Summary of conclusions . . . . .
286 ( 1 ) INTRODUCTION .
THE present work is a continuation of that published in ' Phil. Trans. , Royal Society , ' vol. 214 , pp. 109-146 , 1914 ( Parts I. and II .
) and vol. 215 , pp. 79-103 , 1915 ( Part III * It will lead to clearness in the following development of the subject if a brief of these papers is give1L At the same time , I wish to discuss one or two points in connexion with the views which have been previously advanced and relation between mechanical and molecular theory .
In Part I. the experimental evidence brought forward has justified the hypothesis of molecular distortion enunciated at the outset .
We have thereby been led to regard the molecular configuration of a material medium as a distorted one , and this applies particularly to a substance which is crystalline .
The extent of this distortion is small , but is sufficient to account for the observed change of specific susceptibility which occurs on crystalhzation .
Such change will naturally depend upon the particular crystalline symmetry assumed by the substance .
The theoretical treatment given in Part II .
is an attempt to account for the phenomena observed by extending the electron theory developed by LANGEVIN so * For brevity , reference to these researches is given under Parts I. , II .
, and III .
[ Published , Apri19 , 1920 .
as to include the effects of mutual molecular influences .
Such a representation of the facts has led to the recognition of a large molecular forcive , in all media , depending upon the nature and proximity of the molecules in any particular crystalline grouping .
The existence of this intrinsic molecular field can only be inferred indirectly .
Although the actual properties of the crystalline state depend upon the operation of this field , yet , except in the case of substances of a ferro-magnetic nature , there is no direct experimenta evidence which discloses its very great magnitude .
From a theoretical point of view , there seems to be no doubt that the mutual actions of the molecules are represented by enormous internal forcives in all crystauin media .
The usual method of determining the forcive at an internal point of the material medium is to take a cavity whose dimensions are small in comparison with ordinary lengths ( e.g. , 1 cm .
) and yet large compared with molecular dimensions .
convenient designation of the dimensions of the cavity is contained in the phrase " " physically small.\ldquo ; * In molecular theory , the subdivision of the medium into elements is not valid beyond the limits of physical smallness and only in media which are absolutely continuous may the ] ements be pushed to limits of " " mathematical smallness In a continuous medium our mathematica !
functions give us an accurate estimate of the forcives and potentials operating at internal points ; in a medium composed of discrete particles these same functions give us only an approximate estimate .
A discussion of the nearness of the approximation which can be obtained for material media is of great importance from the point of view of our subject .
For the liquid state the question has been considered by Sir JOSEPH LARMOR who found that the part contributed to the forcive at any internal point by the molecules immediately surrounding that point was , on account of rapid motions and irregular distributions of the axes of the molecules , negligibly small .
To quote from LABMOR : ' ' The general conclusion may be expressed , in an adaptation of CAUCHY'S terminology , by the principle that whenever the integrals in the formulae for mechanical foroes on a material medium cease to be convergent , their principal values must be substituted and again in the footnote to p. 265 , " " This statement ( i.e. , the above quotation ) may be considered to be the mathematical expression of the principle of the mutual compensation of molecular forcives , for which , cf. , 'Phil .
Trans , 1897 , p. 260 .
The principal value of CAUCHY , as regards the completely defined analytical integrals of Pure Mathematics , would be the value at the centre of a minute spherical cavity .
But the quantities which , to avoid periphrasis , have been here called integrals , are really summations of contributions from finite though very small , and complexly constituted , polarised molecules ; the distribution of these molecules that occupy our minute cavity is entirely LEATHEM , " " olume and Surface Integrals used in sics ' Cambridge Monographs , ' No. 1 , p. 6 .
and Matter , ' , 261 , p. 265 , CONSTITUTION AND TEMPERATURE ON MAGNETIC SUSCEPTIBILITY .
249 unknown and may be continually changing , so that the only possible principal value is the one that omits the contribution of neighbouring molecules altogether When now we come to the case of crystalline media , the molecules which were removed from our small spherical cavity would affect considerably the value of the forcive at the centre , for there is in this case no averaging out on account of random motions and orientations of the molecules originally occupying the cavity .
As we do not know the relative dispositions of the molecules composing the space lattice , or the law of force which is operative between the molecules , it is quite impossible to calculate the value of the intrinsic forcive for a point inside a crystalline medium , and as the method of averages , the application of which is quite satisfactory when the medium is in a fluid state , breaks down in the case of a crystalline medium , it is clear that our only way of progress lies in indirect deduction from experimental facts which record the change of physical properties accompanying the transition from the liquid to the crystalline state .
This is the method which has been adopted in the previous portions of this work and it has been shown that the internal forcive at a point of a crystallin medium is extremely large and comparable , if interpreted magnetically , with the molecular field in ferro-magnetic substances ( gauss ) .
In Part III .
it was shown that the potential energy of the molecules forming a crystalline structure was sufficiently large to account for the magnitude of the thermal energy required for fusion .
On p. 201 of his " " Ether and Matter LARMOR states : " " These various actions ( referring to the disturbance of configuration of steady orbits in molecules by action of an applied magnetic field ) involve energy terms for each individual molecule , and the sum for all the molecules , if it could be formed , would represent the total energy of the disturbance of the medium .
But such a mere aggregate of terms would be of no use for applications to matter in bulk ; what we are concerned with there is the mechanical part of the energy , which must be an analytical function of the specification of matter by volume , determined as to mathematical form by the character of the molecular actions , but with coefficients whose values are to be obtained only by direct experiment Although for a fluid medium the total energy of the disturbance of the medium , due to the application of a magnetic field , has little significance , yet , in the transition from the liquid to the crystalline state , during which the molecular field becomes operative , the sum total of the energy disturbance of the medium due to the action of this molecular field is representative of the latent thermal energy which is absorbed when the crystalline medium is fused , and has a definite value for each particular substance .
From the point of view of fluids , the intrinsic forcives mutually compensate and the mathematical functions may be treated as analytic , their principal values being taken .
In a crystalline structure , however , the functions cannot be treated as analytic .
Indirectly we have obtained a measure of the intrinsic forcive in this case with the aid of experimental data .
The most we can obtain by direct experiment , however , is a measure of the mean molecular field , which expresses mathematically how the susceptibility of matter in bulk depends upon a transition from the liquid to the crystalline state .
When thus considering diamagnetic matter in bulk , the large local forcive which has been shown to bind the molecules of the structure together need not be considered , since for matter in bulk its effects are cut out by the mutual compensation of molecular forcives .
It is only when we enquire into the molecular structure of the crystalline medium , or to changes in this structure , that we pass to the inner limit where the principles of LARMOR and CAUCHY for the fluid state no longer apply .
LARMOR remarks : " " The result of the integration still however gave us a valid estimate of the effect of the material system as a whole , when we bore in mind that the infinite or rather undetermined term entering at the inner limit really represents the part of the result which depends solely upon the local molecular configuration ; a part whose actual magnitude could be determined only when that configuration is exactly assigned or known\ldquo ; ( loc. cit. , p. 125 ) .
It is with this " " infinite or rather undetermined term . .
which depends solely upon the local molecular configuration\ldquo ; that these researches are mainly concerned .
It has been called the local molecular field of the crystalline medium ( Part III .
, p. 83 ) .
( 2 ) ON THE ENERGY AND ULTIMATE TENSILE STRENGTH ASSOCIATED W1TH CRYSTALLINE MEDIA 0R GELS .
The large intrinsic potential energy associated with a crystallin medium has been discussed in para .
5 , Part III .
, pp. 90-95 .
It now remains for us to examine the accompanying stresses to see how far the elastic properties of material media may be interpreted in terms of these intrinsic forcives .
Consider first the case of a liquid which is actually cooled in liquid air so.that it passes into a -hard transparent gel when it arrives at the temperature of the liquid air .
It has been suggested ( Part III .
, p. 81 ) that the appearance of rigidity in the gel is due to an interlocking of the irregularly shaped molecules ( arranged at random ) whose thermal agitation is sufficiently reduced .
On account of this random orientation of the interlocked molecules the gel will be isotropic .
At , such a low temperature , also , the molecular motions will be highly constrained so that a particular molecule will present practically the same aspect to the surrounding molecules over a long period .
If this is the case , then.the local molecular forcive between this and a neighbouring molecule will act in a definite direction and will not be rapidly changing its direction as would be the case with the same molecules at a considerably higher temperature ( in the ordinary liquid stats ) .
It is clear therefore that between the of the gel at low temperature we shall have a large local forcive in operation , due tu the interaction of the magnetic systems or revolving electrons within each molecule , but the direction of the action of this forcive between any pair of molecules will be one of random distribution , as we pass from.pair to pair of molecules , although at any given point it is fixed in direction .
CONSTITUTION AND TBMPERATURE ON MAGNETIC SUSCEPTIBILITY .
On further cooling , the molecules continue to readjust themselves and the rigidity increases until a glass is formed .
Still lower temperatures , accompanied by further molecular readjustment due to reduction of amplitude , may result in such a distortion of the lines of force binding the pairs of molecules together , that a completely new pairing of molecules takes place , resulting in spontaneous crystallization accompanied by thermal evolution as the more stable crystalline state is formed .
Whatever may be the nature of the forces which hold the molecules of a liquid together , we have , in addition to these forces , the intrinsic field referred to above , when the substance passes into a rigid gel or crystallizes , and it is due to this intrinsic field that the two latter media show rigidity .
If be this intrinsic field , the local intensity of magnetization , the potential energy per unit volume associated with the gel or crystalline medium will be . . . . . . .
( 1 ) and this will be over and above any potential energy which the molecules of the liquid possess .
This is also a measure of the mechanical stress which binds the molecules of the gel or crystalline medium together and determines their rigidity .
In Part III .
we have given reasons for locating the source of the local molecular field within the molecule and we found that in the immediate neighbourhood of a Fig. 1 .
molecule the value of this field , as determined from the properties of crystalline media , is of the order gauss .
However the molecule is orientated , provided that orientation is not variable with time , the local forcive will be of this order of intensity and in some direction determined by the orientations of the two molecules between which it acts .
In a gel , as we pass from molecule to molecule , the direction of this stress will be continually changing ( fig. 1 ) .
Throughout a crystal , on the other hand , its direction will be constant and will in fact be one.of the determining factors of a particular form of crystalline symmetry fig. .
In a gel , the whole collection of molecules is bound together into one homogeneous *See , p. 86 . .
cit. , p. 90 .
isotropic mass ( i.e. , as viewed in bulk ) whereas in a crystalline medium the mass be aeolotropic .
In Part IIL evidence was brought forward showing that in the came of diamagnetic media the local intensity of magnetization I is of the o1der 100 , while the local molecular field between the molecules is of the order gauss .
The energy of the molecular con{iguration of the crystalline medium over and above that due to Fig. lt the molecular configuration in the liquid state ) or of the gel at low temperatures , will therefore be of the order .
ergs per unit and the internal stress dynes per square centimetre , or 1000 atmospheres approximately .
For wrought iron the energy per unit volume will be ergs and the internal stress about 5500 atmospheres .
In nickel the intrinsic pressure is dynes per square centimetre or 1400 atmospheres ; in cobalt dynes per square oentimetre or 4400 atmospheres , in cast iron 4 dynes per square centimetre or 4000 atmospheres .
These internal stresses are .
measure of the forces bindin the molecules together and should give an estimate of the ultimate tensile strength or tenacity of the medium .
Moreover the tenacities of ferro-magnetic , paramagnetic and diamagnetic media should be roughly of the same order .
That this is so is seen from the following Tenacity Material .
( dynes per square centimetre ) .
, ( wire ) . . . .
Ferro-magnetic Nickel ( wire)Iron , wroughtIron , 5 .
' Mild steel ( perHigh carbon steel cent. carbon ' ( Nickel steel ( 5 per oent .
) .
* This , as we have seen in Part IIL , p. 93 , is compatible with the values of the latent heat of fusion of diamagnetic crystalline media .
See also infra , pp. 2 3-4 , p. KAYE and LABY , ' Physical and Chemical Constants , ' 1918 , p. 28 .
CONSTITUTJON AND TEMPffltATURE ON MAGNETIC SUSCEPTIBILITY .
Material .
Tenacity ( dynes per square centimetre ) .
Diamagnetic . .
Quartz . . . . .
10 Copper . . . . . .
Silver wire ) . . . . .
Gold ( wire ) . . . .
Paramagnetic .
We may conclude that the intrinsic stress due to the local molecular field , calculated as above , gives a satisfactory interpretation of the ultimate tensile strength of crystalline media , as observed experimentally , irrespective of the nature of their magnetic property .
The energy associated with the medium , in virtue of this intrinsic pressure of crystallization , is consistent to the right order with the value of the latent heat of fusion of the materia The energies per cubic centimetre of various diamagnetic and paramagnetic media , calculated from the latent heats of fusion , are given below:\mdash ; 1 .
ORGANIC Compounds Investigated in Parts I. and III .
Xylene Chlorobenzene .
Bromobenzene .
Aniline Acetophenone Benzophenone .
Phenylhydrazino Pyridine Nitrobenzene Naphthalene .
Naphthylamine .
Acetic acid * See also Part III .
, p. 93 .
2 .
ELEMENTS and Inorganic Compounds .
In the case of iron at the point , the transformatiou is from one cubic crystalline form to another , and we should expect the change of internal energy to be smaller than in the general case of actual crystallization from the liquid state .
We can obtain a measure of intrinsic pressures in cgstalline media in another way , which depends on extrapolation of the relation connecting the temperature of the freezing point with applied pressure .
If volume of 1 gramme of liquid , volume of 1 gramme of crystal , 3temperature of fusion , pressure , in atmospheres , latent heat , we know that If the applied pressure be such that then , and if we can determine the pressure for which this condition exists , we have determined the intrinsic pressure due to the crystalline grouping , for if the latter were greater than , or less than , there would be a change of volume on crystallization .
If since , as the curve of fusion shows , both and are finite , must be zero , i.e. , we have to * A. E. OXLEY , ' Trans. Faraday Sooiety , ' vol. XI .
, Part 2 , February , 1916 .
lzUTION AND IEMPEEATURE ON MAGNETIC SUSCEPTIBILITY .
determine the value of the applied pressure corresponding to the of the closed area on the diagram .
We cannot expect this method to give us anything but an approximate value of because ffl extrapolati beyond laboratory pressures is considerable , but the results of the calculation are suggestive .
Tffi relations between the temperature of fusion ( 3 ) and applied pressure ( p ) in atmospheres , for the substances*here referred to are taken from 'Kristallisieren und Schmelzen ' by G. TAMMANN , Leipzig , 1903 , p. 204 , et seq. For water TAMMANN found Differentiating and equating to zero , or 5000 atmospheres dynes/ square centimetre .
Benzophenone\mdash ; Differentiating and equating to zero we find the intrinsic pressure 10,000 atmospheres .
Acetophenone\mdash ; 7700 atmospheres .
Aniline\mdash ; 9000 nospheres .
robenzene\mdash ; 10,000 atmospheres .
Xylene\mdash ; 10,000 atmospheres .
Benzene\mdash ; 7100 atmospheres .
* Most of these substances show a change of diamagnetic susceptibility on crystallization .
See Part I. , pp. 120-131 ; Part III .
, pp. 96-97 .
VOL. CCXX.\mdash ; A. 2 Naphthalene\mdash ; 10 , 100 atmospheres .
tetra.-chloride\mdash ; 11,900 atmospheres .
Ethylene di-bromide\mdash ; 10,000 atmospheres .
Formic acid\mdash ; 7980 atmospheres .
Potassium\mdash ; 10,000 atmospheres .
Phosphorus\mdash ; 27,500 atmospheres .
Sulphur , rhombic-monoclinic* atmospheres .
Solid 13,300 atmospheres .
These valtles , it is true , appear rather high when compared with those found ffom other considerations , but due importance must be attached to the difficulties of experimental work of this nature and to fact that the experimental data have .
been extrapolated over a considerable pressure interval ( several thousands of atmospheres ) .
Comparing these results with those given on pp. 252 to 254 it is considered that a mean value of the intrinsic stress in diamagnetic crystalline media , viz. , 2 dynes per square centimetre is representative of the true order of magnitude of the forcive which binds the molecules in the space lattice of a crystalline medium .
This implies that the energy per unit volume of the diamagnetic crystalhne medium , in virtue of the crystalline grouping , is comparable with 2 ergs .
*This transition of sulphur from the rhombic to the monoclinic form is accompanied by thermal absorption .
Since is positive the transition line for different pressures will be convex to the paxis .
CONSTITUTION AND TEMPERATURE ON MAGNETIC SUSCEPTIBILITY .
As the corresponding stresses and energies are of the same order in ferromagnetic metals,.and , further , since it has been shown ( see Part II .
, pp. 143 , 145 and Part III .
, pp. 84-87 ) that the local molecular forcive in diamagnetic media is of the same order as that in the ferro-magnetic metals , we may conclude that the local intensities of magnetization in the two types of media are comparable .
Since I ergs , and uss ( see Part III .
, p. 86 ) , we find I400 , and In a diamagnetic crystalline medium the local forcives are comparable with in iron , and , since the latter medium shows hysteresis in a magnetic field , we may uire whether a similar phenomenon will be shown by diamagnetic media .
If the diamagnetic molecules are magnetically unsymmetrical , the application of an external magnetic field will tend to orientate them .
* But this will be a differential effect on our conception of a diamagnetic molecule , and thus the tendency of the applied field to produce new molecular groupings will be small .
We should therefore expect that hysteresis due to magnetization will be inappreciable in diamagnetic media .
In iron , on account of the unbalanced magnetic nature of the molecules or atoms , new are actually .produced under fields of moderate intensity and the formation of these implies a loss of energy which is measured by the area of the hysteresis loop .
If , however , we take a diamagnetic copper wire and subject it to mechanical strain , the medium shows mechanical hysteresis .
If sufficiently large stresses are employed , a permanent set is produced within the individual crystalline grains , new groupings of the molecules are formed , and a certain amount of energy is dissipated .
All media , whether they are ferro- , para- , or diamagnetic , will show mechanical hysteresis .
The difference from a magnetic point of view lies merely in the compensated nature of the diamagnetic molecule as compared with the uncompensated nature of the ferromagnetic molecule , but the local forcives are comparable , so that under mechanical stress the mechanical hysteresis effects will be comparable .
( 3 ) A COMPARISON 0F THE ELASTICITIES 0F S0ME DIAMAGNETIC CRYSTALS W1TH TH0SE 0F CRYSTALLINE PARAMAGNETIC AND FERRO-MAGNET1C MEDIA .
It well known that the application of an external magnetic field alters the distribution of stress in a mass of iron crystals .
On our theory we see how the *This orientation in a diamagnetic liquid gives rise to the induced magnetic double rofraction .
See .
Part III .
, p. 87 .
Z58 MAJOR A. .
E. moleculeq orientate ' themselves during crystaUizatiora under the th knal forcives which are characteristic of the mokcular configuration .
The of them local forcives wffi produce in the crystalline medium a distribution of lnternal s which will in : general be different acro different planes , and in way the ]anoe of cleavage can be defined , In the direction where the stress is oatest , krld .
expect the elastic properties of the crystal to be abnormally high , comparable with the elastic properties of steel .
In other directions we should expect the elastic properties to be less pronounoed , and indeed the shearing of crystals , merely by the insertion of a knife blade and the .
application of small pressure parallel to a plane of cleavage , is evidence of this .
The following values of YOUNG'S Modulus of Rigidity for various paramagnetic , and diamagnetic media show that the ) to resist distortion is of the same order whatever the magnetic nature of the crystalline medium .
* KAYE and LABY .
' Physical and Chemical Constants , ' p. 27 , 1918 .
A. E. H. LOVE , The Mathematiaeal Thoory of C$tnb .
Univ. Poess , ' p. .
In the cases of beryl and topaz the fforent values corroepo1ld to bars whose lmgtk are in of [ he different axes of symmetry .
CONSTITUTION AND Ti ; MPERATURR ON MAGNETW LITY .
Z59 As points the values of these elastic constants for beryl and topaz are remarkable in that they are greater than the corresponding constants in ordinary The values of the elastic coefficients for most of the other substances in the above table are comparable with the constants for steel , and it is considered that .
these results give very strong evidence in favour of the large intermolecular forcive operative in diamagnetic crystalline media and confirm the suggestion made in Part II .
, p. 143 , that this local forcive is comparable with that in.ferro-magnetic media .
a crystalline medium be heated , then as long as the crystallme state prevails , rotational vibrations of large amplitude are prevented , so that the specific heat of the crystalline medium is lower than that of the supercooled liquid .
* In the latter case , the liquid at low temperatures passes into a rigid gel , and when this is heated , the molecules acquire rotational vibrations gradually until finally the ordinary liquid state is reached , possessing no appreciable rigidity .
It is important to note that the molecules are vibrating under a local forcive to which we are ascribing the elastic properties of the medium , and therefore the theory is consistent with the theory of specific heat developed by DEBYE , in which the forces which control the thermal vibrations of the molecules are identical with those which determine the elastic constants of the medium .
MADELUNG and SUTHERLAND have similarly suggested that the elastic forces resisting mechanical strain are just those forces which determine the infra-red optical vibrations of the atoms in the solid substance .
It has been found possible to calculate the infra-red frequencies from a knowledge of the mechanical properties .
In the present researches it has been shown that we can calculate both the optical and the mechanical stresses from the local molecular forcive .
Within the core of the atom the local controlling force may be more intense , and although such an intense forcive would not be directly operative in determining the state of crystallization , yet it might be responsible for determining frequencies on the ultra-violet side comparable with -ray frequencies .
( See infra , pp. 273 and 278 .
) ( 4 ) THE CHAN 0F DENSITY 0N CRYSTALLIZATION INTERPRETED As A MAGNETO- STR1CT10N EFFECT 0F THE MOLECULAR FIELD .
If we subject a liquid to a magnetic field , a change of volume occurs to such an extent that the internal pressure is reduced by an amount equal to the potential energy per unit volume of the magnetic field .
This change of internal pressure ( see Part III .
, p. ) . . . . .
( 2 ) where is the susceptibility of the Iiquid per unit volume , , a constant equal to 1/ 3 , and , the applied field intensity .
*Part III .
, p. 94 .
A. .
E. OXLEY , .
' Roy .
Soc. Proc , vol. 95 , p. 58 , 1918 , and Part III .
, p. 84 .
MAJOR A. E. OXLEY ON THE INFLUENCE OF MOLEOULA As for diamagnetic media is of the order , the second term in is insignificant in comparison with If be the compressibility of the liquid , the change of volume in cubic centimetres per cubic centimetre will be given by . . . . .
( 3 ) This relation has been experimentally verified by QUINCKE .
* The compressibility is of the order , particular values for different substances being:\mdash ; Substance .
Benzene . . . . .
Chlorobenzene . . . . .
Toluene . . . . .
Xylene Water Carbon tetrachloride . . . .
Acetic acid . . . .
Carbon bisulphide . . . . .
Substance .
Mercury Potassium .
Sodium . . . . .
Lead . . . . .
Tin . . . . .
Bismuth . . . . .
Iron . . . . . . . . .
Since the largest magnetic field at our disposal is 50,000 gauss , th largest value of is - Now we have shown ( Part III .
, p. 90 ) that the potential energy term corresponding to ( 2 ) for a crystalline medium is . .
per unit volume , where is the constant of the local molecular field and I is the aggregate of .
the local intensity of magnetization per unit volume .
The term is associated with each cubic centimetre of the crystalline structure whether * See G. T. WALKER , " " Aberration and the Electromagnetio ield ' Camb .
Univ. pp. 72-83 .
CONSTITUTION AND TEMPERATURE ON MAGNETIC SUSCEPTIBILITY .
be zero or not .
It is really due to the spontaneous local intensity of magnetization per unit volume and corresponds to the similar energy term , ' ( WEISS ) , due to the spontaneous magnetization in iron .
Using the values which have already been assigned to and , . . . . .
( 4 ) I400 we see at once that this term is large in comparison with and therefore the change of volume to which the.potential energy term gives rise will be large in companson with that which we can produce artificially in a liquid by applying the largest field available in the laboratory .
When a diamagnetic substance crystallizes , the alteration of internal pressure will be and therefore the accompanying change of volume in cubic centimetres per cubic centimetre will be and substituting from ( 4 ) we find to the appropriate order .
Some values of in cubic centimetres per cubic centimetre Substance .
Benzene . . . .
] .
Naphthalene Benzophenone .
Formic acid -phenylamine Sodium . . . .
Potassium . . .
Mercury . . .
LeadTin . .
S Bismuth Iron ( at point ) The values calculated agree as well as could be expected with the experimental determinations , since we know the orders of magnitude only of and I , for these are * See p. 257 supra .
A physical explanation of this large value of and of the corresponding constant in ferromagnetism is given on p. 267 infra .
G. TAMMANN , ' Kristallisieren .
Schmslzen , ' Leipzig , 1903 , pp. 204 et seq. S DESQII , ' Metallography , ' p. 242 .
functions of the molecular It be noticsd that those substances showing a small value of have low bihties .
The case of iron at the transformation is particularly nteIPA .
Fzun ffie curves given by CHARPY and GRENET , *dealin with the expansion of iron and ironcarbon alloys between C. and C. , we can show that the extent of linear contraction which occurs suddenly at the point ( 900 C. ) is of the order centimetre per centimetre .
The change of volume will be of this order of magnitude , which is small in comparison with the change of volume accompanying the tion of many organic compounds , but is very large compared with the magnetostriction effect which can be induced in either a ferro-magnetic or diamagnetic substance with a field of 50,000 gauss .
Taking into account the small of iron , which.is only , or about 1/ 200 that of the liquids above referred to , this change of volume may be interpreted as due to a change of internal energy represented by .
N. where is the constant of the ferro-magnetic field , of the order the intenrty of magnetization , of the order 1760 .
For we have . .
which is of the order of magnitude found experimentally .
The molecular field exists in an unmagnetized of .
iron and is accompanied by the large spontaneous magnetization of that element throughout an individual grain , but as these grains have all types of orientation , the large molecular field and the accompanying spontaneous magnetization are hidden in a piece of iron large enough to contain many grains .
The molecular field will nevertheless produce the 1nagneto-striction effect referred to above .
S Let us suppose that such a piece of iron issubjected to an external magnetic field .
.The molecules of all the grains will tend to come into alignment with the applied field and there will be a new distribution of stress .
In an unmagnetized piece of iron , taken as a whole , the stress may be regarded as equal in all directions , but when an external field is applied , this is no longer the case and the iron shows a new magneto-striction effect consisting of an expansion in one direction and a contraction in the other .
The extent of the redistribution of stress should be deter- mined by a term of the form .
where is the constant of the ldecular field .
cit. RICHARDS , ' Journ. Chem. Soc vol. 99 , p. 1201 , 1911 .
WFJSS and Journ. de Phyg ser. , vol. 7 , p. 249 , 1908 .
S See infra , p. 2.66 .
It is assumed here that the molecular field disappears just above \amp ; Are poin4 at least in so far as it is effective in causing spontanoous tiza .
This is in .
with small paramagnetic susceptibility of above CONSTITUTION AND TEMPERATURE ON MAGNETIC SUSCEPTIBILITY .
263 and is the resultant intensity of magnetization induced in the direction of the external field .
The compressibilities will now be different along and perpendicular to this direction , and the change of volume should be proportional to for the particular value of the applied field .
This result agrees with experiment .
* We can see in a general way how the sign of on crystallization may sometimes be positive and sometimes negative .
Usually the molecular packing in the crystalline state will be closer than that in the liquid state , but it may happen that the configuration of the molecule is such that , when the parts which have the strongest magnetic attraction for each other are in the position of minimum potential energy , the packing is more open than in the liquid state .
The appearance of the internal forcive on crystallization will in this case be panied by expansion .
In the former case the pearance of the internal forcive will be accompanied by contraction .
The region of stability of the crystalline state is represented on the pressure temperature diagram by a closed area ( fig. 2 ) .
Fig. 2 .
This area will in general be divided into four quadrants by the loci of the lines and , where is the change of volume on crystallization and is the latent heat .
In the four quadrants the following conditions hold : If the melting-point is at some point the arc AB of the first quadrant , increase of pressure raises the m.p. and and are This corresponds * NAGAOKA and HONDA , ' Phil. Mag , p. V0L .
CCXX .
Therefore which gives a shrinkage in iron on heating through the point at about C. Fig. 3 .
Fig. 4 .
The application of the relation is interesting in connexion with some abnormalities of heats of recalescenoe in .
ferromagnetic media .
The change of volume , both expansion and contraction ; the absorption or evolution of heat of the amount , depend solely upon the shape of the region of crystalline stability and its position relatively to the axes .
Thus nickel steel with no carbon and pure cobalt show no at the magnetic change points .
In these cases is very small and since and are finite , be large ( fig. 5 ) .
In other words , under the pressure of one atmosphere , the path of the crystallization curve cuts the line AB at a steep angle the neighbourhood of the intersection of the fusion curve with the neutral line CONSTITUTION AND TEMPERATUIffl ON MAQNETIC SUSCEPTIBILITY .
The apparent discontinuities of the susceptibility temperature curve are suggestive in this connexion ( fig. 6 ) .
The upper branch of the curve AB for the so-called -iron is practically continuous with the branch CD for the -range above C. Between and there is a break , the branch BE representing -iron over a somewhat narrow range just below the critical temperature and the branch representing paramagnetic -iron .
The locus BEFC corresponds to a crystalline modification of iron which is more stable over this temperature interval than the crystalline grouping or groupings over the ranges AB and If this view is correct , the molecular field is operative in iron over a temperature interval from C. upwards as well as below the critical temperature .
The existence of this forcive above C. implies a crystallin symmetry involving an appreciable mutual action between the molecules consistent with the enhanced susceptibility found by CURIE* and by WEISS and FOEX over this range .
In the Fig. 5 .
Fig. 6 .
intermediate range , between C. and 1400o C. , iron shows a paramagnetic quality only .
Perhaps we may regard the molecular state in this range as more allied to a gel , consisting of very small interlocked grains , each with relatively few molecules , rather than to a coarse grain crystalline arrangement of the molecules .
The orientations of the molecular axes as we pass from one small grain to another will be different , so that each grain is , as it were , surrounded by a surface of vitreous material .
As stated on p. 251 the molecular field would then be non-effective in so far as the production of * Annals de Chimie et de Physique , ' ser. iv .
, vol. , p. , 1895 .
'Archives des Sciences , Geneve , ' ser. 4 , vol. xxxi .
, p. 88 , 1911 .
This smallness of grain structure above in wrought iron or mild steel is consistent with the experiments of J. E. STEAD , ' Iron and Steel Institute , ' 1898 , No. 1 , p. 145 .
The reverse effect , that of a coarse structure on cooling from above , observed by STBAD and CARPENTER Iron and Steel Institute , ' No. 11 , p. 119 , 1913 ) in the case of thin strips of electrolytic iron , may possibly be attributed to surface forces .
See also a paper ' ' On the Part Played by the Amorphous Phase in the Hardening of Steels by J. C. W. HUMFREY , Trans. Faraday Soo May , 1915 : spontaneous magnetization is concerned , but the tenacity might even tk$ of the chaotic -state since the transition is aocompanied by .
Ainka .
Suob increase of tenacity was actually The effects of prolonged heating on the plasticity of mild steel are this connexion .
and have shown that the plastioity of amaterial is caused by slips occurring on cleavage or " " gliding\ldquo ; surfaces within eaeh of the crystalline grains , although the elementary portions of .
the erystals their primitive form and the crystalline structure of the metal as a whole is preservul In the case of mild steel , exposure to a temperature of 1200o C. or higher temperatures for several hours may cause the material to lose much of its plasticity , while some specimens of soft iron after prolonged exposure at C. to C. ( lees than the critical temperature ) have been made brittle .
These results are consistent with a similarity of molecular configuration for temperatures above C. and below the critical temperature ( about On the other hand an exposure at C. or 1000o C. ( i.e. , in the region of the paramagnetic state ) , followed by a slow or fairly rapid cooling , induces considerable plasticity in the material , and this treatment may even be used to remove brittleness originating from heating to the higher or lower ranges of temperature mentioned above .
As the plasticity is pmduced by slipping on cleavage surfaces within the crystalline grains , this smallness of the grain structure may , under stress , determine a molecular rotation .
EWING and ROSHiNHAIN have shown that in some metals , in addition to slips or motions of pure translation , there results a molecular rotation from strain which gives rise to twin-crystals .
It is interesting to note that the formation of twin-crystals is common in iron through the -range , but has not been observed in the and ranges .
As the temperature falk below 1400o C. there is some modification of the crystalline cubic arrangement , resulting in a closer packing of the molecules , and accompanied by an interlocking of the fine grains .
Thus iron in the -range ( 1400o C. to C. ) will be paramagnetic .
At a lower temperature than 90 C. , this state is unstable and another modification of the crystalline grouping occurs , accompanied by expansion , thermal evolution , and the appearance of spontaneous magnetization .
This latter effect seems inconsistent with a more open packing of the mokcules , but an analogy is found in the case of water , where the molecular influence in the liquid just above C. is small compared with that in ice just below freezing point , although the packing of the.molecules in the liquid state is closer than that in the crystalline state ( see supra p. 264 ) .
As the temperature is lowered the transformation progresses rapidly until a point is reached , after which the increase of magnetic quality is somewhat less rapid .
On continued cooling , the iron passes into the -range where the magnetic property is capable of attainin a saturation value .
' Roy .
Soc. Proc , vol. ; 'Iron Steel Institute , ' No. 1 .
1913 .
' Phil. Trans. Roy .
A. vol. See EWING , ' The Strength of rials , , 1906 .
CONSTITUTION AN TEMPERATURE ON MAGNETIC SUSCEPTIBILITY .
1N FERRO-MAGNET1C AND DIAMAGNETIC MEDIA .
This subject has already been discussed in Part III .
, pp. 89 and 100 , but no interpretation was then given to the magnitudes of the constants and of the local ferro-magnetic and diamagnetic fields .
The values of given by WEISS and BECK* are for iron , and for nickel .
For diamagnetic cryst , alline media is of the order EWING and Low have shown that in very strong magnetic fields the relation between induction ( B ) and applied field ( H ) may be represented by the equation constant . . . . . . .
( 1 ) This constant has the value ( where I is the saturation intensity ) and is equal to 21,360 in wrought iron , 6470 in nickel and 16,300 in cobalt .
In the case of wrought iron . . . . . .
( 2 ) Suppose we could apply a field equal to the molecular field , gauss for iron .
The limiting value of the permeability for this field will be from ( 2 ) ' and the limiting susceptibility per unit volume and ' .
This is equal to the value of , the coefficient of the molecular field , .
as we should expect .
Similar calculations may be made for nickel and cobalt , the limiting susceptibilities being respectively for nickel , .
for cobalt .
Now we may ask the question , why is it that , in spite of the fact that all the molecules are ordered into a definite space lattice under the influence of the respective molecular fields , the materials still show a finite susceptibility to magnetization ?
The * 'Journ .
de Phys ser. , vol. 7 , p. 249 , 1908 .
'Phil .
Trans. Roy .
Soc , 1889 .
explanation , I think , is to be found in the finite though small angular tions which constitute a portion of the thermal energy of the mokculea The molecules are fixed relative to one another and form a definite space lsttice but they are oscillating with small amplitude under the molecular field .
This aUows them to retain a finite susceptibility .
Suppose we could double the molecular field , the limiting susceptibility would become one-half its value and the saturation intensity of magnetization would be slightly increased .
If we could increase the molecular indefinitely , the susceptibility would get indefinitely small , the product of the two however tending to a finite limit equal to the true saturation intensity of magnetization .
The amplitude of the molecular oscillations would , under the influence of this indefinitely large forcive , be indefinitely small .
This state might be attained in a practical manner by cooling the substance , say in liquid hydrogen , when the limiting susceptibility would become vanishingly small .
As Nis the reciprocal of the limiting susceptibility the constant of the molecular field will become indefinitely large .
WEISS , however , supposes to be constant .
* The tendency of to approach a small limiting value as the temperature is lowered is confirmed experimentally for ferro-magnetic is particularly noticeable in the case of weak magnetic fields .
The reduction of the amplitude of vibration of the molecules as the absolute zero is approached merely implies a higher frequency of angular oscillation under the increasing molecular field and does not necessarily imply that the rotational energy becomes vanishingly small .
In this case it should be noted that the saturation intensity of magnetization we are considering is smaller than that which would be given by the simple summation of all the magnetic moments of the molecules in unit volume .
In other words , the difficulty of oducing this latter saturation by an external field becomes increasingly difficult on account of the larger molecular forcive at low temperatures , in agreement with the vanishingly small susceptibility referred to above .
At higher temperatures the susceptibility to an external field is far greater ; the molecules are , as it were , helped over their difficulties with reSpeCt t0 the field , When the eXternal field , by the inCreaSed energy of the rotational osciHations , and having passed this critical point they are held in new combinations .
Beyond the critical point the molecular state is chaotic , the molecules being interlocked ( cf. p. 265 ) , and the external field has sufficient control to produce a paramagnetic effect only .
* Following EISS , we have taken the molecular field proportional to I. WEISS writes the molecular field NI and assumes to be constant .
This applies with sufficient accuracy in a temperature region just ' below the critical temperature , but cannot be true over the region down to absolute zero , because , as the molecular translational vibrations die down , the molecules approach one another more closely and the molecular field must necessarily increase rably although I remains practically constant .
This increase is accounted for by the increase of the coefficient , which is the of the limiting susceptibility .
EWING , ' Magnetic Induction in Iron and other Metals , ' seq. , where curves are given for iron , hard steel , nickel and various nickel steels .
also infra and EWING , , p. 354 .
CONSTITUTION AND TEMPERATURE ON MAGNETIC SUSCEPTIBILITY .
The fall off in the value of as the temperature is reduced , may , in part , be explained by the increased value of the molecular field , due to the nearer approach of the molecules .
As the local molecular field becomes very big , the induced diamagnetic effect in each molecule will become big , in the same proportion , and this will tend to reduce the value of and make this quantity tend to a limiting value .
If the field is of the order gauss at ordinary temperatures , we have seen that the ratio of the induced diamagnetic moment to the magnetic moment of the electron orbit , is of the order 1/ 100 .
If the molecular field at low temperatures approaches 10 ' gauss , owing to closer proximity of the magnetic elements in neighbourin molecules , the diamagnetic effect would be comparable with the ferro-magnetic effec On our view this does not imply that diamagnetic substances should acquire a large diamagnetic susceptibility at very low temperatures .
For taking the molecules in pairs , locally they are paramagnetic and the action of the local molecular field is to reduce this paramagnetic effect so that the local magnetic moment becomes pmaller and the susceptibility to an external field tends to zero as in iron .
According to eriments carried out to test this effect have there proved nor disproved this theory , probably because the external fields were not sufficiently strong .
But during crystallization we are applying unconsciously to each molecular current a magnetic field 500 or 1000 times stronger than the largest field we can apply externally , and probably even greater local intensities are attained at low temperatures , sinoe the interacting magnetic elements in adjacent molecules.may be almost touching .
one another .
The mutual induction and temperature effects combine to cause X to approach the limit zero at the absolute zero or in very powerful external fields .
In diamagnetic media we have seen that the constant of the local molecular field ( which corresponds to Nin WEISS'S ferro-magnetic field ) is of the order and the reciprocal of this , , is the order of magnitude of the local positive susceptibility of a diamagnetic crystalline medium .
( At ordinary temperatures the diamagnetic susceptibility per unit volume is of the order ) .
The parts of molecules adjacent to one another in a diamagnetic crystalline medium attract in similar manner to the adjacent parts of molecules of a ferro-magnetic or paramagnetic medium .
In a fnet medium , as the temperature is raised , the susceptibility increases up to a certain point just below the critical temperature and then falls off rapidly .
The temperature controls susceptibility in two ways ; first , by the molecules to overcome the difficulties of orientation , produced by the neighbouring molecules , to a point just below the critical temperature ; second , by overdoing this and by the molecules too much rotational energy , at the critical temperature and above , so that the susceptibility to magnetization falls very rapidly .
* In nickel the molecular field is 6 gauss .
An applied field of gauss would make a substance as diamagnetic as bismuth have a saturation value equal to that of nickel .
' Magnetic Induction in Iron and other Metah , ' p. 363 .
In a diamagnetic crystalline medium , as the .
temperature is raised , the loaal poiitive susceptibility will obey a similar law , the rature of fusion now corresponding to the critical temperature in the ferro-magnetic case .
locally the relation between susceptibility and temperature is the same in the two the effict unnoticed in the diamagnetic case because the molecule has a total aero magnetic moment .
Nevertheless , the effect of temperature acts in its two antagonistic ways in diamagnetic as well as in ferro-magnetic media .
When the temperature is above the melting point , the rotational energy of the molecules annuls the local forcive ( liquid state ) , when it is very low the molecules become interlocked and cannot readjust themselves in a space lattice ( gel state ) .
There is an intermediate region of temperature where opportunity is offered for the tendency of self-orientation under the mutual local forcives to display itself , and over this range crystallization may take place .
This intermediate temperature range defines the closed region of stability of the crystallin form on the pressure temperature diagram of equilibrium of the crystalline and amorphous states ( see p. 263 ) .
A discussion as to how far we may regard it as proved that the local molecular field in crystallin media is of magnetic nature was given in para .
8 , of Part IIL The conclusion reached was that the molecular field is certainly in part magnetic .
It is possible to bring forward further evidence of the truth of this deduction .
In some noteworthy researches*published by , as long ago as 1870 , it was shown that magnetic properties of crystalline media bear a close relation to molecular aggregation .
About 100 different crystals were examined and from the deportment of these , when subjected to a magnetic field , TYNDALL found that " " if the arrangement of the component particles of any body be such as to present different degrees of proximity in different directions , then the line of closest proximity , other circumstances being equal , will be that chosen by the respective forces for the exhibition of their greatest energy .
If the mass be magnetic this lin will stand axial , if diamagnetic , equatorial.\ldquo ; The exactness of the dependence of magnetic deportment on the position of cleavage planes is remarkably shown in these experiments .
Whatever the crystal examined , it was found that the magnetic deportment disclosed accurate information of the planes of cleavage .
TYNDALL describes the results of his important ' , experiments in such elegant language that it may be permissible to quote some of them at length .
Thus he continues:\mdash ; ' ' From this point of view , the deportment of the two classes of crystals represented by Iceland spar and carbonate of iron , presents no difficulty .
This crystalline form is the same , and as to the arrangement of the particles , what is true of one will be true of the other .
Supposing then , the line of closest proximity to coincide with the optic ; this line , according to the principle expressed , will stand axial or equatorial , according as the mass is magnetic or diamagnetic , which is 'On Diamagnetism and Magnecrystallic Action , ' Loc .
cit. , p. 23 .
CONSTITUTION AND TEMPERATURE ON MAGNETIC SUSCEPTIBILITY .
precisely what the experiments with these crystals exhibit .
Analogy as we have seen justifies the assumption here made .
It will , however , be of interest to enquire , whether any discoverable circumstance connected with crystalline structure exists upon which the difference of proximity depends and knowing which , we can pronounce with tolerable oertainty , as to the position which the crystal will take up in the magnetic field .
" " The following experiments will perhaps suggest a reply .
" " If a prism of sulphate of magnesia be suspended between the poles with its axis horizontal , on exciting the magnet the axis will take up the equatorial position .
This is not entirely due to the form of the crystal ; for even when its axial dimension is shortest , the axis will assert the equatorial position , thus behaving like a magnetic body , setting its longest dimension from pole to pole .
" " Suspended from its end with its axis vertical , the prism will take up a determinate oblique position .
When the crystal has come to rest , let that line through the mass which stands exactly equatorial be.carefully marked .
Lay a knife-edge along this line , and press it in the direction of the axis .
The crystal will split before the pressure , disclosing shining surfaces of cleavage .
This is the only cleavage the crystal possesses and it stands equatorial .
Sulphate of zinc is of the same form as sulphate of magnesia , and its cleavage is discoverable by a process exactly similar to that just described .
Both crystals set their planes of cleavage equatorial .
Both are diamagnetic .
" " Let us now examine a magnetic crystal of similar form .
Sulphate of nickel is , perhaps , as good an example as we can choose .
Suspended in the magnetic field with its axis horizontal , on exciting the magnet the axis will set itself from pole to pole , and this position will be persisted in , even when the axial dimension is shortest .
Suspended from its end , the crystalline prism will take up an oblique position with considerable energy .
When the crystal thus suspended has come to rest , mark the line along its end which stands .
Let a knife edge be laid along this line and pressed in a direction parallel to the axis of the prism .
The crystal will yield before the edge and discover a perfectly clean plane of cleavage .
" " These facts are suggestive .
The crystals here experimented with are of the same outward form ; each has but one cleavage , and the position of this cleavage with regard to the form of the crystal , is the same in all .
The magnetic force , however , at once discovers a difference of action .
The of the magnetic specimens stand .
of the magnetic , cvcial .
" " A cube cut from a prism of scapolite , the axis of the prism being perpendicular to two of the parallel faces of the cube , suspended in the magnetic field , sets itself with the axis of the prism from pole to pole .
" " A cube of beryl of the same dimensions with the axis of the prism from which it is taken also perpendicular to two of the faces , suspended as in the former case , sets itself with the axis equatorial .
Both these crystals are magnetic .
VOL. CCXX .
" " The former experiments showed a dissimilarity of action between diamagnetic crystals .
In the present instances , both are magnetic , but there is $ difference ; the axis of the one prism stands axial , the axis of the other equatorial With regard to the explanation of this .
the following fact is significant Scapolite cleaves parallet to its axis , while beryl cleaves perpendicular to its axis ; the cleavages in both cases , therefore , stand axial , thus agreeing with sulphate of nickel The cleavages hence appear to take up a determinate position regardless of outward form , and they seem to exercise a ruling power over the deportment of the crystal " " A cube of saltpetre , suspended with the crystallographic axis horizontaL sets itself between the poles with this axis equatorial .
" " A cube of topaz , suspended with the crystallographic axis horizontal , sets itself with this axis from pole to pole .
' ' We have here a kind of complementary case to the former .
Both these crystak are diamagnetic .
Saltpetre cleaves parallel to its axis ; topaz perpendicular to its axis .
The planes of cleavage , therefore , stand in both cases equatorial , thus agreeing with sulphate of zinc and sulphate of magnesia .
" " Where do these facts point ?
A moment 's speculation will perhaps be allowed us here .
May we not suppose these crystals to be composed of layers indefinitely thin , laid side by side , within the range of cohesion , which holds them together , but yet not in absolute contact ?
This seems to be no strained idea ; .for expansion and contraction by heat and cold compel us to assume that the particles of matter in general do not touch each other ; that there are unfilled spaces between them .
In such crystals as we have described these spaces may be considered as alternating with the plates which compose the crystal .
From this point of view it seems very not , ural that the magnetic laminae should set themselves axial , and the diamagnetic equatorial .
" " Our mental idea is , that crystals of one cleavage are made up of plates indefinitely thin , separated by spaces indefinitely narrow .
If , however , we suppose two existing at right angles to each other , then we must relinquish the notion of plates and substitute that of little parallel bars ; for the plates are divided into such by the second cleavage .
If we further suppose these bars to be intersected by a cleavage at right angles to their length , then the component crystak wilt be little cubes , as in the case of rock-salt and others .
By ihus increasing the cleavages , the original plates may be subdivided indeffiitely , the shape of the little component crystal bearing special relation to the position of the planes .
It is an inference which follows immediately from our way of viewing the subject , that if the crystal have several planes of cleavage , but all parallel to the same straight line , this line , in the case of magnetic crystals , will stand axial ; in the case of diamagnetic , equatoriaL It also follows that in the so-called regular crystals , in rock-salt , for instance , the cleavages annul each other , and consequently , no directive power will be exhibited , which is actually the case The above quotation from TYNDALL'S work clearly shows how closely allied are the CONSTITUTION AND TEMPE ATURO ON MAGNETIC SUSCEPTIBILITY .
different forces of crystallization in different directions ( which forces determine the planes of ) with the magnetic behaviour of the crystallized medium and lead us to suspect that the forces of cohesion are probably of magnetic nature .
The fine points are so completply explained by the magnetic deportment that it is difficult to dissociate the crystalline forces from a magnetic origin .
If we assume that these foroes are of an electrostatic nature , then it must be admitted that the electrostatic axis of the molecule must coincide with the magnetic axis if the action of a magnetic field is to be decisive , as TNDALL proved it to be , in isolating the planes of cleavage .
But if the electrostatic and magnetic symmetries of the molecules are coincident the application of a field of either nature should induce a double refraction of the same kind in a given liquid .
This , however , is not true experimentally , the electric induced double refraction in liquid carbon bisulphide being opposite in sign to the magnetic induced double refraction .
* Moreover , in crystallin media , the greatest axes of the ellipsoids representing the magnetic and electric properties of the molecule do not in general coincide .
We may therefore say that the evidence points to the conclusion that the forcive which holds the molecules together in a crystallin space lattice is magnetic in nature and not electrostatic .
DRUDE , in his experiments on the relation between valency and dispersion , *COTTON and MOUTON , 'Comptes Rendus , ' vol. 155 , p. 1232 , December , 1912 .
[ Note added Aprd 26 , 1919.\mdash ; After the present communication had passed out of my hands , an important paper " " On the Origin of Spectral Series\ldquo ; was published by Sir J. J. THOMSON ( ' Phil. Mag April , 1919 ) .
In this a new theory of atomic structure is suggested in which the atomic nucleus and the revolving .
electrons play similar roles to those described on p. 274 .
Within the contour of the atom , according to Prof. TIIOMSON , the electrostatic forcive due to the nucleus is of a periodic character and determines a series of spherical or approximately spherical surfaces where the electric force vanishes and over which the periodic motion of the boundary electrons is determined solely by the magnetic field of the atom .
This magnetic field is supposed to be radial .
If this is the case , these intra-atomic fields must be of the order of magnitude gauss simple calculation shows , since to account for the frequencies of the visible spectrum .
Still larger intra-atomic fields will exist nearer to the nucleus , of the order gauss .
These will be sufficient to account for the frequencies of the series .
The infra-red series will be accounted for by fields of the order gauss .
But this latter value is of the order of the intermolecular magnetic field which has been deduced independently in various ways in the present researches .
Moreover , it is to this local field that we have ascribed the rigidity and other properties of crystalline media in general .
The frequencies of the infra-red series will , on this view , correspond with the elastic vibrations of the rigid medium in conformity with the quantum theory of specific heats of EINSTEIN and DEBYE as already stated .
Reasons have already been given for assigning a magnetic nature to the intermolecular field in crystalline media ( see Part III .
, pp. 101-3 , and surra , pp. 270-276 ) .
This intermolecular magnetic field , which is of the order gauss , is suggestive in connexion with Prof. THOMSON'S theory , referred to above .
On p. 274 ( footnote ) it was suggested that the forces determining cohcsion are magnetic in nature , the symmetry of the magnetic forces being determined , however , by the electrostatic action of the nucleus .
Therefore , this fundamental sense , the esent theory and that of Sir J. J. THOMSON are identical .
] 'Ann .
der Phys vol. 14 , and p. 936 , 1904 .
suggested that the electron couples constituting the molecule were of two : ( 1 ) those of the atoms themselves , the sum of which presumably determine the atomic weight ; ( 2 ) those of valency which alone are sufficiently free to vibrate gynchronously with light waves and hence are particularly concerned in the reffaction and dispersion of light .
The valency or boundary electrons are vibrating under the control of the nucleus , but are less firmly held in the system than those near the nucleus .
These valency electrons which have periods corresponding with luminous vibrations are affected by an external magnetic field in accordance with the well-known Zeeman and diamagnetic effects .
The highly constitutive nature of the magnetic susceptibility is consistent with this view that the origin of the magnetic property is partly located near the molecular boundary .
The nuclear electrons in the free atom will determine symmetry of the molecule , and are directly responsible , by their magnetic effect , for the symmetry of the crystallin grouping .
This latter will therefore be determined by the nucleus , which controls the nuclear and boundary electrons , in an indirect manner and the distribution of atomic nuclei in accordance with crystalline symmetry as disclosed by -ray methods is apparent .
W. H. and W. L. BRAG have shown the difficulty , even in simple cases , of definin the molecular boundaries in a crystalline space lattice , although in some cases this is possible .
But to determine by the -ray method whether in any given crystal any atom has a special relation to a neighbourin atom would be practically impossible .
The -ray effects which they investigate are determined only by the nucleus or core of the atom and the outer electrons of the atom which contribute to its magnetic property , though they are controlled by the nucleus , -are probably distorted by the influences of neighbouring magnetic elements .
This distortion , which explains a large number of observed phenomena , defines the molecular boundary within the space lattice and determines a definite chemical molecule .
These outer regions of the atom or molecule remain undetected by the -ray experiments .
* * A. E. OXLEY , ' Nature , ' No. 4 , 1915 .
The core or electrostatic part of the atom is at a much greater distance from the atomic boundary than are the circular currents which give rise to the magnetic properties .
As the intensity of the magnetic field due to a circular current varies inversely as the cube of the distance , and as in a crystalline structure two such circuits may approach so as almost to each electron describing a snmll circle , the local magnetic forcive may be sufficiently large to account for the facts .
The view that the cohesive force in crystalline media is of a magnetic nature was expressed in Part II .
of this research , pp. 83-86 .
It was there stated that in a diamagnetic crystalline medium the molecules are held together by the local magnetic forces due to the revolving electrons .
It is possible that each electron is completely bound to its own nucleus by a narrow tube of force , when the molecules would be capable of attracting or repelling one another electromagnetically according to their directions of rotation , The advantage of an electromagnetic cohesive force lies in the fact that by it we can readily see how similar molecules will cling together .
Electrostatically such attraction implies an electron transfw which , we know , does not always take place .
Reasons are given on p. 277 that the atomic forces which determine ture of olecule qre in part at least agneti .
: CONSTITUTION AND TEMPFRATURE ON MAGNETIC SUSCEPTIBILITY .
The theory propounded by was called by him the " " theory of reciprocal induction and the direction within the crystalline medium where the molecules had the closest proximity and along which the greatest energy was displayed , he called the " " line of elective polarity This theory is identical with our hypothesis of mutual molecular distortion enunciated at the beginning of Part I. and subsequently confirmed in a variety of ways by other physical phenomena .
The direction of closest approach of the molecules , i.e. , the lin of elective polarity is the line along which the crystal shows the maximum elastic properties .
TYNDALL'S explanations of the phenomena he had discovered were prophetic .
The diamagnetic forces were known to be so minute that the theory of reciprocal induction appeared incredible , and , as a correspondence between Lord KELVIN and TYNDALL shows , former expressed emphatically his view that this theory was quite incapable of accounting for the effects observed .
On our modern conception of the magnetic structure of matter , this doubt is dispelled and the smallness of the diamagnetic property is no barrier to the theory of reciprocal induction .
The effect of applying pressure to a diamagnetic medium , produced , in the direction of the pressure , an increase in the diamagnetic property .
This was attributed by to the mutual actions of the diamagnetic polarities which are so minute that their effects , as then understood , would be of such a small order of magnitude that they could not be detected by experiment .
On our view of a diamagnetic molecule , which maintains that such a molecule is paramagnetic locally , the effects observed by TYNDALL can be accounted for quantitatively for the local molecular forcives are comparable with those in para- and ferro-magnetic media .
But even in the case of ferro-magnetic media it is not obvious that the magnetic forces are sufficient to explain the mechanical phenomena unless we realise the localised nature of the forcive .
If we take , for example , a crevasse of the usual assigning a magnetic nature to the forces oi valency is clear , for in this way , without admitting an electron transfer between the various atoms forming the molecule , we can secure the necessary attraction , and this by a fixed or directed force which at the same time is compatible with a characteristic orbital frequency such as appears to be necessary to account for ordinary absorption , magnetic rotation , and diamagnetic phenomena .
The possibility of a satisfactory interpretation of many problems suggested by stereochemistry , in terms of the magnetic force due to revolving electrons , has been ably expounded by A. L. PARSON ( ' Smithsonian Miscellaneous Collections , ' vol. 65 , No. 11 , a paper to which , on account of war service , I have only recently had access ) .
Though PARSON'S theory involves new difficulties in connexion with the distribution of positive electricity in the atom , the advantages from a chemical standpoint which he secures by the inlroduction of magnetic forces of chemical combination cannot be denied .
Granting this , it is natural to suppose that bhe cohesive forces , which hold the molecules in position in a space lattice , will be residual magnetic forces , and that they will closely resemble , in distribution at least , the atomic forces determining the configuration of the molecule .
It will be of great interest to see how far such magnetic cohesive forces are capable of interpreting the spacing of molecules in a crystalline lattice in with the distribution disclosed by -ray analysis .
Magnecrystallic action , as we have seen , is explicable in this way .
* Various ' On Diamagnetism and Action , ' conventional dimensions , within an iron crystal where the saturation intensity is I , the mechanical stress is , which is far smaller than the ultimate tensile strength of the iron .
Indeed , EWING* remarks " " we may , if we please , regard the magnetic molecules as pulling at one another across any imaginary interfaoe , while the stmas with which they pull is balanced by thrust in the framework of the iron , but neither the pull nor the thrust is competent to explain the mechanical strains The above value of the stress , viz. , is obtained by taking a crevasse whose gap , although ' ' physically small is sufficiently wide to accommodate several molecules in line .
If we take a narrower crevasse , approaching " " mathematical smallness\ldquo ; in width of gap , we obtain a measure of the forcive acting between the molecules , and this includes the localised forcive NI or which is of the order gauss .
The localised stress across this interface is or , which we have seen to be of the order 2 dynes per square centimetre .
This is of the same order as the ultimate tensile strength of crystalline media both ferro-magnetic and diamagnetic .
The magnetic resistance of joints is interesting in connexion with the localised nature of the molecular field in iron .
It has been shown by Sir J. J. THONSON and H. F. NEWALL that the susceptibility of an iron bar is much reduced if it is severed and the two parts put in contact .
Later , Sir JAMES EWING and W. Low investigated this effect in a more exhaustive manner when the joints were carefully trued up and also for rough joints , under varying pressures .
They found that for a carefully planed joint a compressive stress of 226 kilogrammes per square centimetre restored almost completely the loss of magnetic property produced by cutting , but that this stress had only a small restorative effect in the case of a rough joint .
In the latter case , we may suppose that the number of points of contact between the two parts of the bar is small , in the former that the two portions are in contact over a large percentage of the available area of contact .
Under a compressive stress of 226 kilogrammes per square centimetre , it appears that in the trued-up specimens the order of contact of the molecules is the same as in the uncut metal and therefore this stress is a measure of the internal stress within the material .
As 226 kilogrammes per square centimetre is equal to dynes per square centimetre , this stress , although lower , is comparable with that calculated on p. 252 , and we may regard the width of the resulting crevasse as approaching mathematical smallness , the spheres of influenoe of the molecules on either side of it pping to an extent comparable with the overlap in the interior of the uncut bar ( see also Part III .
, p. 89 ) .
But even with the most carefully faced junction there will be irregularities , coarse compared with molecular dimensions , and in such regions the localised nature of the molecular field will determine a finite air gap which would account for the difference of stress mentioned above .
Perfectly faced surfaces of soft iron or mild steel ( annealed ) might be 'Magnetic Induction in Iron and other Metals , ' p. 254 , 'Proc .
Camb .
Phil. Soc 'Phil .
Mag September , 1888 , CONSTITUTION AND TEMPERATURE ON SUSCEPTIBILITY .
expected , under the influence of an external magnetic field , to form a perfect junction , in other words to become welded.together .
TAYLOR JONES* has obtained an induction as high as 74,200 Maxwells in soft iron under strong fields .
The tension necessary to pull the surfaces apart in this case will ' 7 .
be ' or dynes/ square centimetre , which is about a twenty-fifth of the tensile strength of the material .
( 6 ) ON A MAGNETIC THEORY or CHEMICAL COMBINATION .
On the theory of chemical action developed by Sir J. J. THOMSON , the determining feature of an atom from the point of view of chemical combine is the number of positive valency electrons it possesses .
These electrons are dragged from their loose attachment to nucleus , during chemical combination , and pass from one atom to another .
The two originally neutral atoms thus become oppositely charged and so attract one another and form , as it were , an electric doublet .
Let us look at this problem from the magnetic standpoint .
Each electron orbit is equivalent to a small magnetic doublet and it is interesting to enquire how far the magnetic forces of such doublets may represent the force of chemical affinity .
Recent work on radio-activity , the wide deflections of -rays , and the diffraction of -rays , all poin to a localisation of the electrostatic charges in a minute core or nucleus .
Round this nucleus , and under its control , the valency electrons ( in part responsible for the magnetic properties ) rotate .
It is conceivable , therefore , that the magnetic forces , in addition to the important role they play in crystallization , may also in part be responsible for the forces of chemical affinity .
If , during chemical combination , there is a definite transfer of valency electrons from one atom to another , we should expect to find an abrupt change in the magnetic behaviour of an atom before and after chemical combination .
If , on the other hand , there is no such electron transference , we might expect that the atoms would preserve their magnetic properties , which would be more or less of an additive nature .
In a remarkable series of investigations , PASCALS has shown that in a very large number of organic compounds , the molecular susceptibility is the sum of the atomic susceptibilities of the component atoms , provided the molecule contains no peculiarity of molecular configuration\mdash ; such , for instance , as the ethylene linkage , unsaturated atom , or complex nucleus .
Thus if is the molecular susceptibility , and the atomic susceptibility of a component atom , we have XN where the summation extends to all the atoms in the molecule and is a positive or * Phil. Mag vol. , p. 165 , 1896 .
' The Corpuscular Theory of Matter , ' 1907 , Chap. VI .
See footnote p. 274 .
Also W. M. HICKS , ' Roy .
Soc. Proc , vol. 90 , p. 356 .
S 'Ann .
de Chim .
et do Physique , ' ser. 8 , vol. p. 5 , 1910 .
negative constant for a certain type of peculiarity of molecular configuration .
In normally saturated compounds .
The elements carbon , hydrogen , chlorine , bromine and iodine have constant atomic susceptibilities in a large variety of simple and complex organic compounds .
This suggests that the origin of the valencies of these elements is also the origin of a definite amount of diamagnetism , under the influence of a definite magnetic field .
In other words , a hydrogen atom , in whatever organic compound it is found , has a constant atomic susceptibility equal to while the carbon atom has a constant atomic susceptibility equal to , and so on .
This result of PASCAL'S , in conjunction with\mdash ; ( 1 ) The enormous magnitude of the local molecular field in diamagnetic media , and ( 2 ) The conception of diamagnetism as due to an induction effect in oppositely spinning electrons ( as developed in Parts I. , II .
and III led me to pect that the magneton may be a constituent of the hydrogen molecule .
The calculation showed* that if there is one electron in each hydrogen atom whose period is equal to .
Soc. Proc , vol. 95 , p. 58 , 1918 .
At the time this paper was written , I was out of touch with the latest available data eoncerning the values of AVAGADRO'S constant ( N ) and the ratio .
The calculation was to determine from the relation Taking molecular susceptibility of hydrogen , ( PASCAL ) , ( MILLIKAN ) , e.m.u. , the number of electrons per molecule , sec. , the period of revolution for the line we find on calculation l63 for the moment of the magneton .
This gives for , as calculated from , the value cm .
In this connexion it should be pointed out that , on PLANCK'S theory oi quanta of energy , the constant is consistent with the existence of a unit of magnetism .
Assuming , as NICHOLSON and BOIIR have done , that the angular momentum of the electron is an integral multiple of , CHAIJERS showed that the magnetic moment of nhe electron orbit is This gives e.m.u. , which is 5 times the experimental value of the moment of the magneton .
If we leave aside PLANCK'S theory of energy quanta and adopt instead SONNnRFBLD'S theory of quanta of action , LANGEVIN showed that a remarkable relation between and exists .
He found ' when the law of attraction between the nucleus and electron the inverse square .
This gives for tho magnetic moment of the electron orbit e.m.u. , a value nearly equal to the most recent experimental value of the moment of the magneton , e.m.u. CONSTITUTION AND TEMPERATUR ON MAGNETIC SUSCEPTIBILITY .
that of the line , then the molecular diamagnetic susceptibility of hydrogen can be accounted for , and each electron orbit of radius cm .
has a magnetic moment , nearly equal to that of the magneton , The atomic susceptibility of carbon in combination is shown by PASCAL to be , and in connexion with the additive law this value is consistent with the experimental values of the molecular susceptibilities .
The mean experimental value of the atomic susceptibility of diamond is The mean of these values is , which is probably as accurate a value as is available at present .
But this value is almost exactly twice that of the atomic susceptibility of hydrogen .
Probably therefore the atom of carbon contains two magnetons .
As to the period of revolution we have where moment of orbit , electron charge in radius of orbit , period , and this implies that the period for the carbon atom is not equal to that of the line unless for the carbon atom .
The sum of the atomic susceptibilities of the atoms in the group is Experimentally , PASCAL showed that the difference of molecular susceptibility in number of compounds whose constitution differed by this group was The mean of these two values is which is probably very true value .
But this is almost exactly four times the value of the atomic susceptibility of hydrogen , viz. , , and in this combination , , we may say that each hydrogen atom has one magneton and the carbon atom has two magnetons .
An ethylene linkage , according to PASCAL , lowers the diamagnetic molecular susceptibility of the compound by , while two or more such linkages lower it by .
These values are respectively equal to twice and four times the atomic susceptibility of hydrogen .
As the additive law holds in the case of the simpler liquid hydrocarbons , it will hold for all the others which differ only by groups , and therefore , taking the values of into account , any member of the three homologous series * HONDA , ' Ann. der Phys vol. 32 , p. 1044 , 1910 , gives ; OWEN , 'Ann .
der Phys vol. 37 , p. 693 , 1912 , gives V0L .
ccxx .
may be considered as containing respectively magnetons .
Moreover , PASCAL has shown that the magnitude of , is quite independent of the presence of nitrogen , oxygen , or sulphur in the compound , so that , where the additive law holds for such complex molecules , it will hold for the whole series of compounds formed with the group .
PASCAL also investigated the halogens , fluorine , chlorine , bromine and iodine .
The atomic susceptibilities of these elements , as deduced from the molecular susceptibilities of organic compounds in which they are contained , are given in column 2 , the approximate number of magnetons per atom is given in column 3 .
Bromin.ineSubstance .
Bromin.ineSubstance .
Bromin.ineSubstance .
Bromin.ineSubstance .
Bromin.ineSubstance .
Bromin.ineSubstance .
Bromin.ineSubstance .
It must be admitted , however , that these last four results do not warrant an extension of the magneton theory to these elements .
Moreover , although the results for the hydrocarbons mentioned above are very suggestive , yet there remain difficulties , such for example as the values of , which for the benzene ring is equal to , the interpretation of which does not fall into line with the magneton view .
Further difficulties are met with in the cases of nitrogen and oxygen .
Perhaps these difficulties should be expected , since it has been proved that the additive law breaks down for many atoms , ecially as regards the metallic elements .
When , in addition , we take into consideration that the mutual disturbanoes of the electron orbits , in atoms containing a relatively large number of electrons , have been neglected , the agreement is probably as good as could be expected .
Thus in the case of the hydrocarbons considered above , compounds which show .
no electrolytic dissociation and in the formation of which no transfer of electrons takes place from one atom to another on combination , we might expect that the addition of a hydrogen atom or a group would add a definite amount of diamagnetism to the compound .
But in the case of the metak and some other elements , chemical combination may be accompanied by the transference of electrons , i.e. , by a breakdown of the magnetic elements of the atoms .
In such cases the additive law could not hold .
CONSTITUTION AND TEMPERATURE ON MAGNETIC SUSCEPTIBILITY .
Thus iron-carbonyl Fe and nickel-carbonyl Ni are ; potassium ferri-cyanide is paramagnetic , while potassium ferro-cyanide is diamag- netic .
It would appear that in these cases the loss of magnetic property of the iron and nickel atoms is due to a transfer of valency electrons , i.e. , it involves an electric charging up of the atoms .
The behaviour of the oxygen atom in organic compounds , in compounds with chlorine and in metallic oxides , where it acts always as an electronegative element , may possibly be accounted for in the same way .
Free oxygen and ozone are strongly paramagnetic , but no semblance of an additive nature of the magnetic property is found in any of the oxygen compounds .
The appearance of strong magnetism in the Heusler alloys and its disappearance in manganese steek , are similar effects , dependent on the formation of intermetallic chemical compounds accompanied by an electron transfer .
There are many paramagnetic substances which possess molecular magnetic moments comparable with , and in some cases much superior to , those shown by ferromagnetic substances .
The apparent feeble susceptibility they possess is due to the fact that , with the largest magnetic field which can be applied , we can never produce anything like a saturation effect .
In fact , according to WEISS , the molecule of cobal chloride , Co. contains 25 magnetons , while an atom of cobalt below the critical temperature contains 9 magnetons .
We may well ask ourselves\mdash ; what is the nature of the process by which the addition of a diamagnetic substance , H.C1 to cobalt , produces such a large increase in the number of magnetons per molec.ule .
Assumin the work of WEISS holds good , and there is certainly a very considerable amount of evidence in favour of his theory , we can interpret this result either by supposing that the atom of cobalt really contains more than 9 magnetons , S or else that the diamagnetic acid supplies the additional magnetons when it acts on the cobalt to form the chloride .
In either case , it seems that we must admit that a molecule may possess systems of magnetons which , in certain circumstances , are so arranged to counterbalance one another , producing no additional moment of the molecule These magnetons would contribute nothing to the paramagnetic or ferro-magnetic property of a substance and could not be included in WEISS'S theory .
The grouping of these " " latent\ldquo ; magnetons , according to our extended view , would be perturbed by the union of the cobalt atom with the Cl ion , in a manner similar to that by which an external field reveak the spontaneous magnetization in iron ( as interpreted on EWING'S theory ) by orientating groups which were formerly so constituted as to show no magnetic effect externally .
If this is so , then we are only justified in assuming that purely diamagnetic molecules contain groups of magnetons so arranged that the A. E. OXLEY , ' Proc. Camb .
Phil. Soc vol. 16 , p. 102 , 1911 .
J. S. TOWNSEND , ' Phil. Trans. Roy .
Soc , vol. 187 , p. 647 , 1896 .
The specific susceptibility of cobalt chloride is , that of hydrochloric acid S Some of which are self-compensated .
I.e. , of the molecule as a whole .
2 .
molecule possesses no resultant magnetic moment .
For example , while iron below the critical temperature possesses , according to WEISS , 11 magnetons to the atom , the molecule of ferric sulphate possesses 30 , ferric chloride 28 , sodium ferro-pyrophoephate 26 , sodium ferrous oxalate 3 magnetons to the atom , above the critical temperature either 8 or 9 magnetons to the atom , while a molecule of nickel sulphate contains 16 magnetons .
In general the number of magnetons , per molecule of a salt of a ferro-magnetic element , is large compared with the number of magnetons associated with an atom of the pure ferro-magnetic .
The fact that cupric salts are paramagnetic while cuprous ones are diamagnetic is interesting from our point of view .
Although copper is diamagnetic , yet a molecule of cupric sulphate contains 10 magnetons .
It seems as if the large local atomic fields , which have been recognised in diamagnetic and ferro-magnetic molecules , have the power , when the molecules approach so that their fields overlap , to upset the equilibrium of the atoms in combination and redistribute their magnetic elements .
This is easily possible when the great intensity of the local molecular field is borne in mind .
In most cases a diamagnetic molecule , on account of its symmetry , would , under the influence of such a field , remain diamagnetic , but each orbit would be distorted by the field and the susceptibility of the substance would be slightly modified .
( 7 ) ON SOME ANOMALIES THE MAGNETIC ROTATION EFFECT .
Diamagnetic media are in general dextro-gyric .
The only exception is titanium chloride which is laevo-gyric .
Paramagnetic media are sometimes dextro-gyric and sometimes laevo-gyric , while the ferro-magnetic elements , iron , nickel and cobalt are all dextro-gyric .
At present no theory seems capable of accounting for these anomalies and it is therefore interesting to examine to what extent the local molecular field may cause the effects observed .
At one time it was thought that the direction of rotation probably depended merely on the diamagnetic or paramagnetic property of the molecule , but experiment soon disproved.this generalisation .
VOIGT suggested that the production of an intense reverse field , when the external field was applied , would account for the effects , but no physical explanation of a possible origin for this intense reverse field was given .
On the views of diamagnetic ari paramagnetic polarity developed in these researches , the necessary fields demanded by VOIGT are found in the immediate neighbourhood of the molecular boundary .
In diamagnetic liquids the molecules have zero magnetic moments and their axes are distributed at random .
The application of 'Journal de Phys vol. 1 , ser. , p. 974 , 1911 .
For solutions of salts in water it should be noted that VERDET'S constant for the solute is to be regarded as negative if VERDET'S constant the solvent is less than , which is the value of the constant for pure water at C. See P. ZEEMAN , ' Researches in Magneto Optics , ' p. 186 .
CONSTITUTION AND TEMPERATURE ON MAGNETIC SUSCEPTIBILITY .
an external field roduces a minute diamagnetic polarity , which differential effect determines the magnitude and sign of the rotation .
When a polarised ray passes over such molecules , it will be rotated considerably locally , but only to be rotated in the opposite direction to an almost equal extent in a neighbouring molecule or in another part of the same molecule .
This will also be true if the diamagnetic medium is crystalline .
Hence the final rotation will be a relatively small differential effect compatible with the diamagnetic effect of the medium in bulk .
We shall omit the exceptional case of titanium chloride for the present and pass on to consider the ferro-magnetic media iron , nickel and cobalt .
In ferro-magnetic media , on account of the continuity of magnetic induction , there is an enormous reverse local field , and if this acts over regions of the molecule containin magnetically active electrons , a large rotation will be produced , which will not be compensated in neighbouring molecules , when the ferro-magnetic material is saturated .
Hence we should expect a very large rotation to be produced by such media .
This has been confirmed experimentally , in the cases of iron , nickel and cobalt , by KUNDT .
aramagnetic solutions lie in an intermediate category .
The application of external magnetic field causes a certain amount of molecular orientation depending on the temperature .
Such orientation causes a reverse field over a part of the system , but over the neighbouring molecules in combination with the one considered , the field depends upon the difference the reverse local field and applied field , which difference may be positive or negative .
If , in such regions , there are electrons capable.of orientating the polarised beam , the rotation may be dextro- or laevo-gyric , according as the resultant field is opposite to , or in the same direction as , the applied field .
This will not explain the exceptional case of titanium chloride unless this molecule possesses some peculiar dissymmetry , whereby , in spite of its diamagnetic nature , it becomes orientated under the applied field .
Such orientation is the basis of the magnetic double refraction theory developed by LANGEVIN* and confirmed experimentally in some respects by COTTON and ( see Part III .
, p. 87 ) .
An alternative explanation of the behaviour of the titanium chloride molecule may be found in a rotation of the paramagnetic titanium atom relative to the compound molecule .
BECQuEREL has deduced an interesting connecting the magnetic rotation ( r ) with the Zeeman coefficient , the applied field , wave-length and refractive index , . . . . .
( 1 ) where is the velocity of light .
* LANGEVIN , ' Le Radium , ' vol. 7 , p. 251 , 1910 .
'Ann .
de Chim .
et de Phys ser. viii .
, vol. 19 , p. 155 , 1910 .
SCHUSTER , ' Theory of Optics , ' p. 307 .
: We shall assume that this relation applies to all media .
On our theory be the sum of the applied and local fields and will be considered positive if it is in the direction of the external field .
we shall take to be negative , i.e. , the negative electron is always responsible for the Zeeman effect .
In mwnetic media , in general , is positive , is negative , and therefore is negative , or clockwise looked at along the direction of H. The rotation is dextrogyric .
In ferro-magnetic is the resultant of the applied field and a very large reverse molecular field , so that is very large and negative ; is positive .
Hence again is negative and the rotation is dextro-gyric as in diamagnetic media .
In paramagnetic solutions , may be positive or negative over magnetically active atoms , and therefore the sign of may be positive or negative .
Hence some paramagnetic solutions will be dextro- and some laevo-gyric .
In the ferro-magnetic elements the magnitude of the rotation is high .
Thin films of saturated iron show a rotation of the order 260 million times that of carbon bi-sulphide subjected to an external field of one gauss .
To test this we may write for iron , for carbon bisulphide .
where is the reversed molecular field , of the order gauss , gausa The Zeeman coefficients and we shall take to be of the same order , hence :\mdash ; .
( 2 ) For iron* Near $ 'Smithsonian Physical Tables , ' p. 196 .
CONSTITUTION AND TEMPERATURE ON MAGNETIC SUSCEPTIBILITY .
For.carbon bisulphide* Near From ( 2 ) For nickel Near and For Near Taking the reverse molecular field equal to that in iron These values , although rather small , are comparable with the enormous ratios of the rotations of saturated ferro-magnetics to diamagnetics , as actually observed *Loc .
cit. , p. 192 .
Loc .
, p. 196 .
Loc .
, p. 195 .
286 MAJOR A. R OXLEY ON THE INFLUENCE OF MOLmULAB experimentally , and point to the importance of the local molecular forcive in magnetooptic phenomena .
It should be noted that the value of the reverse field we have taken is that of the intermolecular field .
Within the atom the reverse field probably be greater than gauss .
If it is of the order gauss , the above ratios become for iron , for nickel and 700 for cobalt .
Intra-atomic fields of the order gauss are required by to explain the pressure shift of spectral lines and by RITZ in his theory of spectral series ( see ako Part III .
, p. 100 , and supra , p. 273 , footnote ) .
( s ) SUMMARY 0F CONCLUSIONS .
( I. ) The )plications of the local molecular forcive , in diamagnetic , paramagnetic and ferro-magnetic media , have in the present research been extended to interpret the ultimate tensile strength of crystalline and vitreous media .
It has been shown by EWING and ROSENHAIN that the permanent set which occurs prior to breaking is due to slipping along the cleavage planes within the individual crystalline grains .
We should therefore expect that the material would be fractured when the applied mechanical stress is equal to that produced internally by the local molecular forcive .
The internal stress within the material is shown to be of the order 2 dynes per square centimetre which is approximately the mean value of the ultimate tensile strengths of crystalline and vitreous media ( pp. 250-259 ) .
As a consequence of this internal stress , the energy per unit volume will be 2 ergs , and this energy , which is over and above that which exists in the fluid state , should be a measure of the latent heat of fusion per cubic centimetre .
This test which was applied in Part III .
to test the order of the local forcive , has been extended to a variety of organic and inorganic media , including the metals and is found to accord with the experimental values to the right order ( pp. 253-4 ) .
( III .
) Since the forces under which , the molecules vibrate are those to which we ascribe the elastic properties of crystallin media , the results obtained are consistent with the theory of specific heats developed by DEBYE , in which the specific heat is attributed to purely translational vibration , and it has been shown ( Part IIL ) that , near the fusion point , the rotational energy acquired by the molecules will give a measurable departure from this theory which is actually observed experimentaly .
As we should expect , it is found that the elastic constants of a variety of ferromagnetic , diamagnetic and paramagnetic media are of the same order , several diamagnetic and parama.gnetic media even surpassing steel in their power to resist distortion ( pp. 257-259 ) .
( IV .
) Any change of internal pressure will be anied by a change of volume defined by the compressibility of the medium and dependent as to sign upon 'Astrophysical Journal , ' vol. 23 , p. 232 , 1906 ; vol. 35 , p. 268 , 1912 .
' Ann. der Phys vol. 26 , p. 660 , 1908 .
CONSTITUTION AND TEMPERATURE ON MAGNETIC SUSCEPTIBILITY .
peculiarities of the molecular configuration .
It has been shown that the energy change which occurs on crystallization is compatible with a volume change of the same order of magnitude as that accompanying crystallization , and we may therefore interpret the change of volume on crystallization as a magneto-striction effect of the local molecular forcive .
The magneto-striction effect depends on molecular orientation which is proportional to the square of the magnetic forcive ( pp. 259-263 ) .
In Part III .
it was shown that the double refraction of crystalline edia can be interpreted as due to the magnetic double refraction effect of the local molecular forcive which orientates the molecules into a crystalline space lattice .
This effect is also proportional to the square of the magnetic forcive and the two effects mutually support one another .
( V. ) The above results are interesting in connexion with TAMMANN'S theory of the closed region of stability of the crystallin state , as represented on the pressure ternperature diagram .
TAMMANN'S experimental work gives an alternative method of determining intrinsic pressures , but the results are notably higher than those found in other ways .
Possibly this is due to extrapolation over a wide pressure range .
The pressure temperature diagrams showing the fusion curve are instructive in dealing with problems relating to thermal evolution ( or absorption ) , and volume changes and possible interpretations of these peculiarities in the ferro-magnetic elements , iron , nickel and cobalt , have been given ( pp. 263-266 ) .
( VI .
) A physical interpretation has been given of the large values of the coefficients , and , of the molecular fields in ferro-magnetic and diamagnetic crystalline media respectively .
These coefficients are the reciprocals of the limiting local susceptibihties of the media under field strengths equal to the respective molecular fields .
The local susceptibility of a diamagnetic molecule is comparable : with that of a ferro-magnetic molecule and the two vary in the same way with field strength and temperature .
In diamagnetic media , however , magnetic hysteresis will be inappreciable , since the molecule as a whole possesses a zero magnetic moment. .
Nevertheless , mechanical hysteresis in diamagnetic media will be of the same order as in ferro-magnetic media ( pp. 267-270 and p. 257 ) .
( VII .
) From TYNDALL'S experiments on the deportment of paramagnetic and diamagnetic cries in a etic field , the positions of the planes of cleavage can be traced .
These results show that the forces responsible for crystalline symmetry are very probably of a magnetic nature .
If the forces are of an electrostatic nature , then , since an electric field must disclose the same planes of cleavage , the electric and magnetic symmetries must coincide .
This is not the case however .
The magnetic forces are partly due to the valency or boundary electrons whose orbits are controlled by the atomic nuclei .
The nuclei determine the crystalline symmetry , directly , through the medium of the magnetic forces of the electrons .
This conclusion is not at variance with the results of -ray diffraction experiments ; the latter determine only the positions of the diffuse diffracting cores and give no VOL. CCXX . .
288 MAJOR A. E. OXLEY ON THE INFLUENCE OF MOLWULAS .
indication as to the outer regions of the atom which determine the Hence , in spite of the disclosure of the -ray methods , it is maintained that within a crystallin medium the molecules , though distorted , are still essentially inkgrd .
units and that it is possible to imagine a surface enclosing each ( pp. 270-277 ) .
( VIIL ) The smallness of the nucleus disclosed by recent work on radio-activity suggests that in addition to the part played by the magnetic forces in crystallization , these fOrces are in part responsible for chemical combination ( footnote , p. 274 ) .
The theory of chemical combination developed by Sir J. J. THONSON inlplies the transference of electrons from one atom to another whereby the atoms become oppositely charged .
Such a transference of the valency electrons implies a com- .
plete readjustment of the magnetic property and therefore this property could not be of an additive nature .
This is borne out by a large amount of experimental evidence .
If the chemical combination takes place without a transferenoe of the valency electrons from one atom to another , which probably happens in many organic compounds in which electrolytic dissociation does not take place , we might expect the magnetic properties to be more or less of an additive nature .
PASCAL'S work conffims this view ( pp. 277-281 ) .
( IX .
) The author has previously shown that the magneton may be a constituent of the diamagnetic hydrogen molecule .
It appears that this idea may be extended to carbon and the hydrocarbons in general , where the molecular susceptibility can be directly calculated from the atomic susceptibilities of the component atoms .
The results , however , are not so convincing when extended to other elements , but the fact that departures from the additive law occur in such cases leads us to suppose that some disturbin influence has been neglected .
WEISS'S work on salts of the ferromagnetic elements , taken in conjunction with our present conception of diamagnetism , suggests that diamagnetic substances contain magnetons , compensated so as to produce a diamagnetic effect of the medium in bulk .
The foroes of chemical combination may , however , perturb this state , and by rearranging the magnetic elements give rise to a compound possessing more magnetons than are contained in an unbalanced state in a ferro-magnetic element .
Examples of this are given and also of the verse effect which may equally well arise ( pp. 281-282 ) .
( X. ) The principle of the continuity of magnetic induction , as applied to the local molecular forcive , suggests a possible interpretation of known anomalies of the magnetic rotational effect .
In paramagnetic solutions , the dextro-gyric or laevo-gyric property is attributed to a differential effect of the.reversed local field and the applied field over rotationally active electrons , the sign of the effect dependin on the direction of the resultant field acting over these electrons .
The very powerful rotational effect of thin films of iron , nickel and cobalt , in comparison with that shown by carbon bisulphide , has also received an interpretation in terms of the local molecular forcive ( pp. 282-286 ) .
CONSTITUTION AND TEMPERATURE ON SUSCEPTIBILITY .
( XI .
) It is considered that the above conclusions when conjoined with those btained in Parts I. , IL , III .
, which bring the theory into line with the magnetic atom flelds of RITZ and probably suggest an origin of spectral series , amply justify the importance of the magnetic forcive in crystallin and vitreous media .
The magnitude of this local magnetic forcive , first calculated to interpret the change of diamagnetic usceptibility observed on crystallization of a large number of organic compounds experimented upon in Parts I. and III .
, has been found capable of correlating a number of additional physical phenomena of wide difference of origin .
It is hoped to continue these applications to other branches of optics including spectral series and .optical activity .
( XII .
) [ Added February 28 , \mdash ; It has been established that the intermolecular field , in all crystalline media , is of the order gauss .
The electrons , within the free atom , are controlled by electrostatic , and possibly also by magnetic , forces , whose ( origin lies in the core .
When the atoms are grouped into a definite space lattice , the cohesive force between them is of a magnetic nature , and the rigidity of such media is due to the localised mechanical stress , exerted at definite points across the atomic " " surface\ldquo ; ; the electrons revolving in small circles in adjacent atoms ( p. 274 ) .
This mechanical stress , due to the local magnetic forces , is responsible for the change of specific susceptibility , and other properties , on crystallization ; and is balanced by the stress due to the distortion of the internal e]ectrostatic configuration of the atoms .
In this way , a balance is red between the electrostatic and magnetic stresses ; these stresses predominating alternately , as , we pass through the crystallin structure , thus giving rise to a system in equilibrium .
The rotation of electrons in small circles , at definite points near the atomic surface is su.ggestive in connexion with the theory of directed valencies r.equired explain stereo-chemical phenomena ( p. 274 , footnote

