]\gt ; [ 79 ] III .
The Influence of Jlotecular Constitution and Temperature on Magnetic Susceptibility .
Part III .
On the Molecular Fietd in Diamagnetic Substances .
A. E. OXLEY , M.A. , M.Sc .
, Coutts Student , Trinity College , Cambriclge , Mackinnon Student of the Royal Society .
Communicated by Prof. J. J. THOMSON , O.M. , F.R. Received June 24 , \mdash ; Read June 25 , \mdash ; Revised December 2 , 1914 .
CONTENTS .
Page ( 1 ) Introduction . . . . . . . .
79 ( 2 ) On the local molecular polarization and the local molecular forces within diamagnetic fluid and crystalline media . . . . .
80 The mean and local molecular fields of a diamagnetic crystalline . . . . .
81 ( 4 ) On the magnitude of the local molecular field . . . . . . . . . .
84 ( 5 ) On the stresses and energy associated with the molecular field . . . . .
90 ( 6 ) Additional . . . . . . . . . .
96 ( 7 ) A relation between the magnetic double refraction of organic liquids and the change of magnetic susceptibility due to crystallization . . . . .
97 On the nature of the molecular field . . . . .
100 ( 1 ) INTRODUCTION .
IN the present communication an attempt is made to examine to what extent the crystallin state of diamagnetic substances ( which form a very large proportion of the whole number of substances known to us ) involves mutual actions between the constituent molecules and to obtain a measure of the forces which hold the molecules together in a definite space lattice characteristic of the substance .
The work is thus a continuation of that published in 'Roy .
Soc. Phil. Trans vol. 214 , pp. 109-146 , 1914 , wherein it was shown how from determinations dealing with the change of diamagnetic property on crystallization information could be derived concerning the inner structure of crystalline media and the intensity of the interacting molecular forces .
In the paper cited ( p. 143 ) it was suggested that the local molecular field within diamagnetic crystalline media may be comparable with the ferro-magnetic molecular [ Published , 1915 .
MR. A. E. OXLEy ON THE INFLUENCE OF MOLECULAR field .
In the light of the experimental results on aromatic substances there described and of those given by BOIS , HONDA and OWEN , which are concerned with elementary substances , this suggestion has been found to lead to interesting results .
It is clear that we can easily test whether such an intense local molecular field exists ; for if so it must be of import in connection with the transition from the liquid to the crystal ] in state in eneral , and changes of other physical properties which accompany this change of must also depend directly or indirectly upon the molecular field .
The object the present communication is to test the experimental and theoretical work which has already been completed by applying the results obtained to a wider of physical phenomena .
Evidence of a change of specific diamagnetic susceptibility owing to crystallization has been obtained with about forty substances , but with some substances the change , if it exists , is exceedingly small .
Ten additional substances have been investigated , and the results , in conjunction with the other physical properties , are used to extend the ideas concerning the molecular field to diamagnetic crystalline media in general .
PART III .
( 2 ) ON THE LOCAL POLARIZATION THE LOCAL MOLECULAR FORCES W1TH1N FLUID AND CRYSTALLINE MEDIA .
The passage from a purely mechanical theory of the properties of material media to the actual discrete molecular theory , as interpreted electrodynamically , involves an examination of the local polarizations which act in the immediate neighbourhood of the point considered within the medium .
It has been shown by Sir JOSEPH LARMOB* that the forces exerted at such a poin by the surrounding polarized medium can be separated into two parts , a purely local part and a part due to the rest of the medium .
If we neglect the forces due to the former then those due to the latter are derivable from a potential due to the combined volume and surface density distributions of POISSON .
With fluid media we can go further and determine , at least approximately , the forces due to the purely local part .
This is attained by imagining a cavity scooped out round the point and determining the force due to the surface distribution of polarization over its walls .
The removed molecules are now put back in the cavity and the additional force which they contribute is found by applying the method of averages .
The molecules of a fluid are in rapid motion and the force at a point in the interior of the cavity due to these replaced molecules averages out to a small COlTectivep .
It can be shown that the force due to the immediately neighbouring molecules is represented by where is the .
Soc. Phil. Trans , vol. 190 , p. 233 .
1897 ; and LORENTZ , 'Theory of Electrons , ' pp. 137 and 303 .
CONSTITUTION AND RATURE ON MAGNETIC SUSCEPTIBILITY .
polarization due to the applied field and is the coefficient of the corrective polarization\mdash ; the effect of the replaced molecules .
Within a crystalline structure , however , this method of averages applied to the molecules within the cavity is not valid , for each molecule is orientated so as to occupy some definite position with respect to its neighbours .
In such a case the local force , due to a lnolecule which is close to the point under consideration , is uncompensated , and we have to resort to indirect deduction imental facts to obtain a measure of the polarization effect due to the molecules immediately surrounding the point .
It is to this uncompensated effect that we have attributed the rigidity of the crystalline medium .
* In order to avoid the complexity introduced by the elasticity of crystalline media LARIIOR confined his investigations to fluids ; but , as we are attempting to ascribe to these very elastic forces a nature which permits of their interpretation in terms of a local molecular field binding the molecules together , the treatment for crystalline substances is the same in kind as that for fluid substances .
The rigidity of gels at low temperatures is of a different nature and is probably due to an interlocking of the molecules ( arranged at random ) whose thermal agitation is sufficiently reduced .
This is in accordance with the work of who found that at very low telnperatures the power and velocity of crystallization were very small .
( 3 ) THE NIEAN AND LOCAL MOLECULAR IELDS OF A DIAMAGNETIC CRYSTALLINE SUBSTANCE .
In a former author has shown that the change of specific susceptibility , which accompanies the transition from the liquid to the crystalline state for many substances , can be satisfactorily interpreted in terms of a molecular field which operates to an appreciable extent only in the crystalline state .
The nature of this molecular field need not be definitely specified , and it will suffice for our purpose to regard it as a magnetic field limited to such a nitude as to produce in the molecules of the substance a distortion , or polarization , such as is equivalent to that actually produced by the mutual forces of the molecules of the crystalline structure .
S In Part II .
of the work referred to , the mean molecular field was represented magnetically by the term , where is a coefficient ( the constant of the 'Roy .
Soc. Phil. Trans , vol. 214 , p. 143 , 1914 .
VlIETJIAM , ' of Solution , ' p. 44 , and references to TAMMANN'S work there given .
'Roy .
Soc. Phil. , vol. 214 , p. 109 , 1914 .
This paper contains Parts I. and II .
of the prcsent work , and , for brevity , referencr , made to these in what follows .
S A force of electrostatic nature will modify the electron orbits , and therefore the susceptibility , an effect which we may represent magnetically .
If the revolving electrons are controlled by magnets ( as on theory ) , a magnetic field would be the true interpretation of the moleculal field .
Probably both types of field -exist ( vide infra , p. 22 , for a discussion of the nature of the molecular field ) .
VOL. CCXV.\mdash ; A. molecular field ) defining the extent of the polarization of the molecules in the crystalline state and depending upon the definite space lattice which the molecules assume .
where is the applied field , defines the distortion produced in an ' electron orbit by the forces exerted by surrounding molecules when the substance is in the liquid state .
is the polarization ( proportional to the applied field H ) due to this distortion .
As we do not know the actual disposition or the proximity of the molecules in the structure , the difficulty in attempting to calculate is insuperable and we are compelled to try the more indirect method afforded by experiment .
rough estimate of the value of has been obtained from CHAUDIER'S observations on the change of magnetic rotatory power when aniline , benzene and nitrobenzene crystallize , the order of magnitude deduced being at least The corresponding value of for the liquid state , deduced from theoretical considerations , is of the order Thus in the liquid state the mean molecular field is represented by .
The distortion of the molecules in the liquid state of the substance is therefore insignificant compared with that for the crystalline state and , neglecting in comparison with it was shown that , . . . . .
( 1 ) where is the intensity of the applied magnetic field , and are the specific susceptibilities of the crystalline and liquid states respectively .
Thus we may ignore the mutual influences of the molecules of the crystalline state providing we supply a mean molecular field whose magnetic equivalent is .
Equation ( 1 ) shows that the actual change of on crystallization ( a few per cent. with aromatic compounds ) demands that shall be of the order 5 .
; which implies a small and reasonable distortion of the molecules of a liquid due to the surrounding molecules .
It should be noted that this is a maximum value of the distortion for the liquid state .
The value will be smaller if , as is probable , is of a higher order than .
At present no physical interpretation has been put upon the large value of , but we see that it is representative of the local part of the forcive in the crystallin state just as the smaller constant is representative of the local part of the forcive in the liquid state .
On the theory of magnetism developed by LANGEVIN , a diamagnetic molecule contains oppositely spinning systems of electrons which counterbalance one another externally so that the molecule possesses no initial magnetic moment .
When an external magnetic field is applied the period of one system of electrons is lengthened , that of the other system is shortened , and the molecule becomes slightly polarized or distorted .
This small differential effect , which is compatible with the Lorentz * Part II .
, p. 141 .
LARMOR , 'Roy .
Soc. Phil. Trans , vol. 190 , p. 233 , 1897 .
This null initial magnetic moment corresponds to the null electric moment possessed by a molecule of a dielectric which is not subjected to an external electrostatic field .
CONSTITUTION AND TEMPERATURE ON MAGNETIC SUSCEPTIBILITY .
explanation of the Zeeman effect , is the origin of the small negative moment possessed by diamagnetic substances when subjected to a magnetic field .
The smallness of the diamagnetic susceptibility in general is due to the self-compensation of the forces ( produced by the oppositely rotating systems of electrons or any other magnetic system in the molecule ) which takes place at points whose distances from the molecules are large in comparison with molecular dimensions .
But quite close up to the molecule these forces will not compensate and the local force will be comparable with that due to an unbalanced revolving electron or other magnetic element of the molecule .
It is with this local field that the present communication is mainly concerned .
I shall call it the local molecular field of the diamagnetic substance .
In the crystalline state we may regard the space lattice as defined by the local forces exerted between a system of electrons in one molecule upon another system in an adjacent molecule and so on throughout the crystalline structure .
As was shown in Part II .
, p. 142 , the mean molecular field is proportional to the intensity of magnetization of the substance , and we may write . . . . . . . . . .
( 2 ) where is the new constant of the molecular field , of the order the number of molecules per gram .
, the diamagnetic moment induced in a molecule by applying an external magnetic field , and is the density of the substance .
The local molecular field will have the same constant of proportionality but we must replace the Fig. 1 .
diamagnetic moment of the molecule as a whole by the local moment associated with a pair of molecules .
We have no means of measuring directly the value of this latter quantity , but it is probably very large in comparison with , and it was suggested at the end of Part II .
that it may be so large that the local molecular field is comparable with the molecular field in ferro-magnetic substances .
On account of the null initial moment which has been prescribed for the diamagnetic molecule the local molecular field will be of an alternating character , the distance over which it is unidirectional being comparable with the distance between the molecules , as the diagram above shows .
* It should be noted that this molecular field will exist in all crystalline substances , whether they are subjected to an external magnetic field or not .
On the other hand , the mean molecular field is a differential effect and depends on , the diamagnetic moment which is induced by the external field .
MR. A. E. OXLEy ON THE INFLUENCE OF MOLECULAR The small value of possessed by fluid substances implies a mean molecular field much smallel than that associated with a crystalline structure .
But how much larger the mean and local molecular fields of the crystalline state are compared with those of the liquid state has not been ascertained at present .
All that has been shown in Parts I. and II .
is that the minimum value of the mean molecular field in some diamagnetic crystalline media is large compared with that in liquids .
We shall now proceed to obtain an estimate of the intensity of the local molecular field in crystalline media from the small variation of which occurs on crystallization .
( 4 ) ON THE MAGNITUDHi 0F THE LOCAL MOLECULAR FIELD .
If we apply a strong magnetic field to a diamagnetic liquid whose molecules possess dissymmetry , LANGEVIN* has shown that the liquid becomes doubly refracting and that a very small proportion of the unsymmetrical molecules suffer an orientation under action of the field .
The amount of orientation which can be produced by the largest magnetic field obtainable is not sufficient to affect the value of the susceptibility appreciably .
When , however , the liquid crystallizes and all the molecules are similarly orientated , we should expect that the susceptibility will be different in different directions if the molecules are not symmetrical .
But this cannot be the explanation of the change of observed in the experiments of Part I. ( and in those carried out by HONDA and OWEN ) .
For in these experiments the crystals were small and their axes would be distributed at random .
In some cases the crystals were observed to grow in such a way .
The explanation here put forward , and which has already been advanced in Parts I. and II .
, is that each molecule of the substance is distorted when it forms part of a crystalline structure by the local forces due to the neighbouring molecules .
This mutual action between the molecules is necessary to account for the rigidity of the crystalline structure and for the existence of planes of cleavage .
We can therefore find the intensity of the local molecular field by making it of such magnitude as to produce a change of susceptibility of the same order of magnitude as that actually observed in the crystallization experiments .
We may write . . . . .
( 3 ) where is the change of moment produced in an electron orbit of moment by applying a field is the period of the electron , approximately sec. , and the ratio of the charge to the mass of an electron , .
The largest field which 'Le Radium , ' vol. 7 , p. 249 , 1910 .
, .
252 .
LANGEVIN , ' Ann. de Chim .
et de Phys ser. VIII .
, ] .
, p. 96 , 1905 .
CONSTITUTION AND TEMPERATURE ON MAGNETIC SUSCEPTIBILITY .
can be produced in the laboratory is of the order gauss , and therefore the largest moment which we can induce is given by . .
This induced moment is sufficiently large to account for the diamagnetic susceptibility of the substance .
From equation ( 3 ) we have , . . . . .
( 4 ) , ( 5 ) where the subscripts and refer to the liquid and crystalline states espectively .
In passing from the liquid to the crystalline state the alteration of produced by the local molecular field is where ( 6 ) . . . . . .
( 7 ) It should be noted that although alternates as we pass from molecule to molecule of the crystalline structure , still the sign of will evel.ywhere remain the same , for when changes sign M also changes sign , so that every molecule of the crystalline structure suffers a definite distortion due to the action of the local molecular field .
The ambiguity of sign in the above formulae simply implies that the arrangement of molecules may be such that is increased or decreased by the particular kind of packing which the molecules assume and corresponds to a reversal of the sign of .
( See Part I. , p. 133 , and Part II .
, p. 143 .
) From equations ( 6 ) and ( 7 ) , we find , . . .
( 8 ) and therefore from ( 5 ) . . .
( 9 ) Hence , ( 10 ) which gives by ( 4 ) .
( 11 ) , the molecular field within the liquid , is insignificant compared with and is therefore eglected iu equations ( 6 ) and ( .
The electrons which give rise to diamagnetism also give rise to the Zeeman effect , a slight alteration of their periods accountin for both phenomena .
* The order of magnitude of the local molecular field will not be effected by assuming that the electrons in a molecule have a mean natural period , when the substance is in the liquid state , which becomes modified to , when crystallization sets in .
The change of period indicates the order of magnitude of the displacement of an absorption band owing to the transition from the liquid to the crystalline state .
Now , where is the number of electrons per molecule and the number of molecules per gram .
Therefore The change of period is produced by the local molecular field , and therefore Hence .
( 12 ) This equation gives us a value of the local molecular field when we know the extent to which is modified on crystallization .
With all the bstances investigated in Part I. and with many of the elements examined by HONDA and OWEN , amounts to a few per cent. Hence either Taking , we get either gauss or gauss , and in either case is of the order gauss .
We have no experimental evidence at present as to how far an absorption line is displaced when a substance passes from the liquid to the crystalline state , but such * LANGEVIN , ' Ann. de Chim .
et de Pkys vol. 5 , p. 97 , 1905 .
CONSTITUTION AND TEMPERATURE ON ILITY .
evidence would be a direct test of the value of The above deduction is an interesting confirmation of the suggestion made at the end of Part II .
, p. 143 , with regard to the intensity of the local molecular field in diamagnetic crystalline substances .
In a crystalline substance which shows natural double refraction we may regard the molecules as held in position in the crystalline structure by this intense local field , and in that case the double refraction of the medium would be a consequence of the orientation of the molecules due to the operation of this field when the substance crystallizes .
Now the magnetic double refraction induced in a liquid is proportional to the square of the field intensity .
us assume that this law holds good for much fields than the largest we can apply in the laboratory .
If we take gauss as the maximum value of this external field then the magnetic double refraction of the liquid will be represented by , where is a constant independent of the applied field .
If we could apply a field of gauss the magnetic double refraction would be , i. e. , 40,000 times as great .
This is of the order of magnitude of the ratio of the double refraction of quartz to liquid nitrobenzene , the latter being jected to a field of gauss .
Conversely , if we assume that on crystallization there is an internal local molecular field , which if interpreted as a magnetic field is of the order gauss , we can see how under its influence the molecules would become orientated to such an extent as to give rise to a crystal possessin the high natural double refraction of quartz .
All elements and compounds which show on fusion a small percentage change of must possess a molecular field whose local value , if interpreted magnetically , is of the order gauss .
This value is also supported by observations on the artificially induced and natural double refractions of other liquid and crystalline media .
It is known that most uniaxal crystals have a double refraction comparable with that of quartz , and of a much higher order of magnitude than that which has been induced in a liquid by the largest magnetic field at our disposal .
S Thus the abnormally high double refraction of Iceland spar is about 100 times that of quartz .
On the other hand , the double refraction of ice is only about of that of quartz .
* It will be shown later that the mechanical interpretation of the large local molecular field is the internal stress which accounts for the rigidity of a crystalline medium .
Now it has besn shown by ( ' Astrophys .
Journal vol. 35 , p. 268 ) that there is a direct relation bstween the pressure shift of spectral lines and the Zeeman effect .
Hence it is very probable that the shift of an absorption band on crystallization would be determined by the Zeeman effect of This results from both the theory of HAVELOCK , 'Roy .
Soc. Proc , vol. 77 , p. 170 , 1906 , and that of LANGEVIN , 'Le Radium , ' vol. 7 , p. 251 , 1910 .
The is found to hold experimentally , CorroN and MOUTON , ' Ann. de Chim .
et de Phys ser. VIII .
, vol. 19 , p. 155 , 1910 .
There are good reasons for supposing that this is not accurate for such large fields , for a saturation effect must come in .
But the result of assuming that it holds is suggestive , and probably the true order of the eflects .
S note , p. 99 .
But COTTON and MOUTON have shown that the magnetic double refraction induced in water by the largest magnetic field they were able to apply ( 3 .
gauss ) is insignificant compared with that of liquid nitrobenzene subjected to the same field .
In eneral the natural double refraction of a crystalline medium is of a hi gh order compared with the artificial double refraction which we can produce in a liquid using a field of 3 .
gauss .
The intrinsic molecular field , if interpreted magnetically , must therefore be large compared with 3 .
gauss in all crystalline diamagnetic media .
This result will be used later in the extension of our ideas concerning the molecular field to diamagnetic crystalline media in general .
In Part II .
the change of specific susceptibility due to the crystallization was represented by the difference , , of the mean molecular fields of the crystalline and liquid states .
, where is the applied field , defines the distortion produced in an electron orbit by the forces exerted by surrounding molecules when the substance is in the liquid state and subjected to a field H. is the polarization ( proportional to the applied field H ) due to this distortion .
is the constant of the molecular field for the liquid state and is approximately equal to .
Now the molecules of a liquid are moving about rapidly in all directions , and therefore the polarization defined by is an average effect .
In the crystalline state , on the other hand , the molecules are orientated into definite positions with respect to one another , and therefore the polarization due to the mutual influences of the molecules in this case ( as disclosed by applying an external field H ) is large compared with Let be this new polarization .
Then , which is large compared with , is a factor defining the poJarization which results from the complete orientation of the molecules in the crystalline medium .
will also depend upon the proximity of the molecules .
Now reyerting to equation ( 8 ) of Part II .
, the variation of on ] ization may be written if is large pared with .
The large value of implies a correspondingly small value of , i.e. , a small value for the molecular distortion in the liquid state in order to account for a given change of on crystallization .
As in S3 we may write the mean molecular field , which accounts for the change on crystallization , in the form ( diamagnetic moment per unit volume ) .
N. , and the value of , the new constant of the molecular field , is of the order for those substances which show a small percentage change of on crystallization .
Now the local molecular field is of the order .
gauss .
We may write .
I where I is the aggregate of the local intensity of magnetization per unit volume .
, the constant of the local molecular field , will be equal to the constant of CONSTITUTION AND TEMPERATURE ON MAGNETIC SUSCEPTIBILITy .
the mean molecular field , for the relationship of the molecules to one another in the crystallin structure determines their common magnitude .
Hence and I100 .
We can now use these values to form an estimate of the potential energy associated with a diamagnetic crystalline medium in virtue of the local molecular field and the local molecular polarization .
But before passing on to this , it will be interesting to compare the above deductions with regard to the local molecular forcive in diamagnetic crystalline media with those of WEISS which are concerned with the forcive in ferro-magnetic media .
Imagine a spherical cavity , large compared with molecular dimensions but lying wholly within the same crystal of ( 1 ) a diamagnetic ; ( 2 ) a ferro-magnetic medium .
On account of the structure which has been assigned to the diamagnetic molecule , the force at the centre of the cavity ( when there is no external field ) will be zero in case ( 1 ) .
In ( 2 ) the force will be where 1 is the spontaneous intensity of ation .
Now let us move our point of observation from ths centre of the cavity out towards the surface .
When the point approaches within a range comparable with molecular dimensions , in case(2 ) , the forcive increases , and when the point is at a distance from the wall equal to that which separates two molecules of the crystalline structure the forcive is represented by where is the constant of WEISS'S lmolecula field , of the order .
If we move our point of observation in case ( 1 ) ( the diamagnetic medium ) from the centre of the cavity to the surface , then as the point approaches to within a distance comparable with molecular dimensions the local force is due almost entirely to the molecule which is nearest to the point .
This molecule maintains a definite orientation with respect to the point , and when the latter is so close to the molecule as to be almost on its surface the polarization is comparable with the saturation intensity in iron .
This is the interpretation of the large values of and I which are each of order gauss .
As , we may suppose that the large coefficients and determine the enormous magnitude of the forcives quite close to a molecule in the respective crystalline structures .
These local forcives we could hardly hope to calculate directly for we do not know the proximity of the molecules in the structure or the law of force which holds at such a close range .
The local molecular field of a diamagnetic crystalline substance alternates as we pass from molecule to molecule of the structure and therefore localized .
If we could take a crevasse between two molecules of the structure then the induction across it would give us a measure of Similarly in the ferro-magnetic case a crevasse of such small dimensions would give us a measure of WEISS'S field I. If we take a crevasse in the ferro-magnetic medium , which is large compared with molecular dimensions , the force in the gap is I and this is small compared with I. The difference between these forces must be attributed to the localization of the intense fields associated with the iron atom .
We then get continuity of magnetic induction while the intense field is still capable of modifying the structure of a neighbouring molecule .
VOL. CCXV .
The crystallization of a diamagnetic substance may be regarded as accompanied by the production of a spontaneous local intensity of magnetization .
We have up to the present regarded the molecular field as a magnetic field , and a close analogy has been found between the phenomena shown by ferro-magnetic and diamagnetic substances .
The question as to how far we are justified in regarding ths molecular fields as magnetic fields will be discussed in S8 .
( 5 ) ON THE STRESSES AND ENERGY ASSOCIATED W1TH THE MOLECULAR FIELD .
In the previous work evidence has been given which shows that the forces associated with a diamagnetic crystalline structure are exceedingly large , and therefore the potential energy of the crystalline state will be considerable .
If be the local magnetic moment which in conjunction with the local molecular field binds one molecule to another in the crystalline structure , and if all such elementary systems ( each system consisting of the adjacent parts of a pair of molecules which are bound together by the local forces ) are independent , then the energy possessed by 1 .
of the substance in virtue of a particular crystalline grouping may be written Let be the number of molecules per cubic centimetre .
Then where corresponds to the molecular field in ferro-magnetism .
If we put and , we find for the energy associated with 1 .
of the substance .
( 13 ) Here is the constant of the molecular field , I the aggregate of the local intensity of magnetization per unit volume , and the density of the substance .
This is the amount of potential energy which the molecules contained in 1 .
possess in virtue of their grouping , and is additional to the energy possessed by 1 .
of the liquid .
We may treat the crystalline substance as a fluid whose molecules do not influence one another , providin the energy term represented by ( 13 ) is superposed upon any energy which 1 .
of the fluid may possess .
Therefore if the crystalline structure be submitted to a magnetic field , the potential energy associated with 1 .
may be written , ( 14 ) where and are the susceptibility and mass per unit volume of the liquid .
Let us now compare this with the expression given by LARMOR* for the potential energy per 'Roy .
Soc. Proc , vol. 52 , p. 63 , 1892 .
CONSTITUTION AND TEMPERATURE ON MAGNETIC SUSCEPTIBILITY .
gramme of a diamagnetic liquid the molecules of which have a small mutual influence .
If be the susceptibility per unit volume , this is shown to be , ( 15 ) where is a constant approximately equal to ( our ) .
In accordance with our notation we have , where is the number of molecules per cubic centimetre , and the diamagnetic moment induced in each by H. Therefore , instead of ( 15 ) , we may write . . . . . . . . .
( 16 ) Comparing ( 14 ) with ( 16 ) we may identify with and I with .
Further , the applied field is associated with the moment , which it produces , in the same manner as the molecular field is associated with the aggregate of the local intensity of magnetization per unit volume , I , which it produces .
The analogy is complete .
In ( 14 ) we are concerned with the local forces of the crystalline structure which , on account of the relative fixity of the molecules , do not average out ; whereas in ( 16 ) we are concerned with the average forces within the fluid , the large local forces having become averaged out on account of the motions of the molecules roun .
any point .
Prof. LARMOR has pointed out an interesting case in which the term in predominates.* If we suspend a bunch of iron 1lails from the pole of a magnet , we find.that they adhere to each other endwise and repel one another sideways , while non-adjacent nails have no action on one another .
This is analogous to the result disclosed by ( 14 ) for a diamagnetic crystalline substance .
molecule may be considered as possessing two little magnets opposing one another , and these molecules fit together in such a way that the magnets nearest to one another in adjacent molecules help one another .
In the case of a fluid and are small , so that the predominant term in the expression for the energy is , as is ordinarily assumed .
For a crystalline medium taken as a whole the corresponding energy term is , the expression usually taken in measuring It is only when we are inquiring into the local forces binding the molecules together into a crystalline lattice , the energy term of which is , that we get the true diamagnetic analogy with the bunch of iron nails .
The molecules of the diamagnetic structure are held together endwise , so to speak , but we may have attractive iorces in a perpendicular direction depending *Loc .
, p. 64 .
Strictly speaking , to each of these expressions for the energy should be added a Germ proportional to , which is very small .
upon the configuration of the molecule .
These latter forces will , in general , be different from the former , and will give rise to a difference of cohesion which accounts for the greater ease of cleavage of crystals in certain directions .
Here , again , we have further proof of the truth of the hypothesis of molecular distortion in crystalline media .
It is true that if to a liquid we could apply such an intense field that all its molecules are orientated , that liquid would possess a double refraction equal to that of a crystal , but as long as there is no mutual action between the molecules this doubly refracting medium could show no signs of rigidity and no preference for cleavage along certain planes .
Clearly the process at work in the formation of a crystalline structure is that of a binding force ( mutual induction in our case ) between two unbalanced parts of two adjacent molecules .
All magne-crystallic properties can readily be interpreted in terms of such a mutual effect .
It is important to note , however , that unless we recognize the enormous intensity of the local molecular field , which , together with the large local intensity of magnetization in diamagnetic crystals , binds the molecules together and by a mutual induction effect distorts them , we could not account for the rigidity of the crystalline medium , or the extent of its double refraction .
It appears that TYNDALL'S the deportment of diamagnetic crystals when placed in a magnetic field as due to a mutual action between the diamagnetic molecules is sufficient to account qualitatively for the behaviour , but it is difficult to see , on TYNDALL'S view of a simple and very minute diamagnetic polarity , where such large forces as those demanded for crystalline media could have their origin .
On our view a diamagnetic molecule as a hole possesses a small diamagnetic polarity , an induction effect of the applied field , and the force due to it at a point considerably removed from the molecule is small .
in between a pair of molecules the internal forces are unbalanced , and the intensity of the local field is comparable with that in ferro-magnetic substances .
On the other hand , quite to the diamagnetic molecule conceived by TYNDALL , the force binding it to a neighbouring molecule is not intense enough to account for the difference of magnetic property in different directions , as Lord KELVIN pointed out .
The explanation of magne-crystallic action formulated by TYNDALL ( this is the theory of reciprocal molecular induction ) accounts qualitatively for the phenomena but certainly fails from a quantitative point of view .
The present conception of a diamagnetic molecule surmounts the latter difficulty .
] As this question of determining the order of intensity of the local forces and local polarizations within diamagnetic crystalline media is of the greatest importance , any additional proof of the correctness of the values assigned to them is valuable .
Let us therefore try to form some estimate of the magnitude of the term , which [ Nwember 12 , 1914 .
] Cf .
the liquid crystalline state .
TYNDALL , ' On Diamagnetism and Magne-crystallic Action , ' 1870 , p. 69 , CONSTITUTION AND TEMPERATURE ON MAGNETIC SUSCEPTIBILITY .
is large compared with the other term of equation ( 14 ) .
Taking the value of the local molecular field , , we find since , that I , the aggregate of the local intensity of magnetization per unit volume , is of the order 100 .
This is comparable with the saturation intensity of ferro-magnetic substances .
If ergs .
, which is the amount of potential energy associated with 1 .
of the crystallin medium in virtue of its molecular grouping .
The thermal equivalent of this will be of the order gr. calories , which represents the heat energy required to destroy the crystallin structure , i.e. , the latent heat of fusion .
This is of the right order of magnitude for many diamagnetic substances\mdash ; organic compounds and elements .
* The above reasoning applies only to the order of magnitude of the latent heat .
It is obvious that until we know the disposition of the molecules within the crystalline structure the value of is somewhat vague .
But the experimental fact that the latent heat of transformation of iron from the ferro-magnetic to the paramagnetic state is of the same order of magnitude as the latent heat of fusion of many diamagnetic crystalline substances is powerful evidence that the local forces and local polarization which ws have assigned to diamagnetic crystalline structures are enormous , comparable , in fact , as the above and preceding calculations have shown , with the intense forces and polarization of ferro-magnetic substances .
In the crystallin state we must regard the molecules as orientated into positions with respect to their neighbours by these large termolecular forces .
If at the higher temperatures the molecules undergo rotational vibrationfi about their mean positions , then it would be expected that the value of will be somewhat lessened by these vibrations , and we should therefore expect that a small fraction of the energy associated with the grouping would be dissipated as the temperature is raised towards the fusion point .
The effect this would have the variation of the specific * The following values of the latent heat for some diamagnetic substances with we directly concerned taken from ' Recueil des Constantes Physiques , ' Paris , 1913 , pp. 323-4:\mdash ; Benzene . . . . . .
30 Xylene . . .
39 Chlorobenzene 30 Bromobenzene 20 Aniline . . . .
21 phenone . . .
33 Benzophenone . . .
Phenylhydrazine .
36 Pyridine . . . .
22 Nitrobenzene . . .
22 Naphthalene 35 Naphthylamine 22 Acetic acid . . . . . .
44 Carbon tetrachloride 4 Bismuth . . . . . .
13 Cadmium 14 Lead . . . .
: 5 Silver . . .
22 Tin . . . .
14 Zinc . . . .
28 Gallium . . . . .
19 Iron ( ferro-magnetic ) 59 heat with temperature would be to add to the normal variation , expressed by DEBYE'S theory , the following positive term . .
, . . . . .
( 17 ) where 3 is the absolute temperature and the mechanical equivalent of the calorie .
In a former paper the author has shown that a term of this nature is necessary to represent the variation of the specific heat of substances in the neighbourhood of the fusion point .
* A corresponding term explains , on WEISS'S theory , the variation of the specific heat of ferro-magnetic substances in the neighbourhood of the transformation temperature , on the supposition of a ferro-magnetic molecular field of the order gauss .
fact that NEBNST and LINDEMANNS have found experimentally an abnormal increase of the specific heat of diamagnetic substances in addition to the normal variation due to purely translational vibrations , as the fusion point is approached , is additional evidence of the importance of the rotational term ( 17 ) .
DEBYE'S quantum theory of specific heats is concerned with translational vibrations of the molecules only , and , away from the fusion point , it agrees well with experiment .
Incidentally , in order that ( 17 ) may be a measurable fraction of the specific heat , and I must be large , for , from experimental data showing the departure from DEBYE'S theory near the fusion point , the interval of temperature over which the molecules have effective rotational vibrations mounts to several degrees at least , so that the large value of ( 17 ) cannot be attributed solely to a large value of the gradient .
Unless and I have values of the order we have already found for them , it would be impossible to account for the measurable departure of the specific heat near the fusion point from DEBYE'S values .
Only a fraction of the energy term will be dissipated below the fusion point , the major portion disappears at the fusion point and corresponds to the latent heat ( as described above ) .
The departure of the specific heat from the value calculated on theory is important in connection with the quantum theory , for if the latter be valid , the above term , due to the rotation of the molecules , implies that the angular velocities of the molecules go in definite units .
We cannot have the quantum theory holding for translational motion and not for rotational .
The remarkable fact is that the rotation al term ( 17 ) is insignificant except near the fusion point .
This means that away from the fusion point the translational motion of the eCuleS is sufficient to * A. E. OXLEY , ' Proc. Camb .
Phil. Soc vol. XVII .
, p. 450 , 1914 .
WEISS and BECK , ' Journ. do Phys ser. , vol. 7 , p. 249 , 1908 .
[ Added November 12 , 1914 .
] S 'La Theory du Rayonnement et les Quanta , ' Paris , 1912 ; particnlarly p. 272 and the memoirs of NERNST and EINSTEIN .
CONSTITUTION AND TEMPERATURE ON MAGNETIC SUSCEPTIBILITY .
account for the observed specific heat , JEANS , in his " " Report on Radiation and the Quantum Theory published by the Physical Society of London , refers on p. 77 to the necessity of the rotational term , which was pointed out by the author in ' Proc. Camb .
Phil. Soc. , ' vol. XVII p. 450 , 1914 .
JEANS adds : " " The absence of a noticeable contribution to the specific heats is accounted for , on the quantum theory , by supposing that the forces opposing rotational movements of the inside the solid are so large that the corresponding vibrations are of very high frequency , , normally , possess very little ener.gy .
As far as pure theory goes , there is no question that to the terms in the specific heat contemplated by NERNST'S theory there ought to be added an additional term of a form exactly similar to the Einstein term , but having , where is the frequency ( or average frequency ) of the vibrations which depend on the rotations of the atoms .
" " It is worthy of note that sodium and mercury show an increase , beyond that accounted for by the theories we have considered , in the specific heats as the fusion points is approached , when , presumably , the intensity of the forces which the atom from rotating is relaxed , and NERNST and LINDEMANN find that in general the same is true for the substances they have examined Before passing on to further experimental work and the extension of our results to crystalline diamagnetic media in general , it will be convenient to collect the results which have been obtained in the preceding pages .
The work contained in Parts I. and II .
has received full support and been confirmed with regard to the enormous intensity of the local molecular field in about 40 diamagnetic substances which show a measurable change of on crystallization .
Evidence that the magnitude of this field is comparable with that of the ferro-magnetic field has been obtained from the following independent sources:\mdash ; ( 1 .
) The change of susceptibility observed on crystallization demands a local molecular field of this order of intensity .
( 2 .
) The natural double refraction of a crystalline substance as compared with the artificial double refraction which can be induced in a liquid by the strongest magnetic field at our disposal is consistent with the value of the local molecular field implied by ( 1 ) for diamagnetic crystalline media .
( 3 .
) ( 1 ) and ( 2 ) together imply that the aggregate of the local intensity of magnetization per unit volume of a diamagnetic substance is comparable with the saturation intensity of magnetization of a ferro-magnetic substance .
( 4 .
) The above Tesults lead to a correct estimate of the energy ( potential ) associated with the crystalline structure , in virtue of the molecular grouping , as tested by the magnitude of the latent heat .
( 5 .
) Lastly , unless the forces binding the diamagnetic molecules together were of the order of magnitude stated , we should not be able to detect a departure of the experimental value of the specific heat near the fusion point from the value calculated on DEBYE'S theory .
Every substance investigated by NERNST and LINDEMANN discloses such a departure .
The above evidence is sufficient to establish the existence of an intense local molecular field of the order gauss , if interpreted magnetically , in those diamagnetic crystalline substances ( about 40 of which have been investigated ) which show a measurable change of on crystallization .
We shall now pass on to some additional experiments with the object of extending the above conclusions to diamagnetic crystalline media in general .
( 6 ) OTTON and MOUrON have found that aromatic liquids show an abnormally large double refraction compared with aliphatic liquids when subjected to the same external magnetic field .
According to the theory of molecular orientation , which ( in the opinion of these authors ) is unique in accounting for all the observed phenomena of induced double refraction , the extent of the double refraction is directly proportional to the degree of dissymlnetry of the molecule .
Now assuming this to be so , we should expect that an unsymlnetrical molecule , whose electrons are more readily displaced in one direction than in another , would have a distortion produced in it , when subjected to the local field of a neighbouring molecule , this distortion being characterized by the molecule 's own dissymmetry .
Therefore those liquids which show the larger induced double refraction when acted on by a magnetic field should also be the ones which show a large value of on crystallization .
All the aromatic liquids examined in Part I. show an appreciable change of on crystallization and , according to COTTON and MOUTON , all these show an easily measurable magnetic double refraction .
With regard to aliphatic compounds , COTTON and MOUTON found that liquid hexane , chloroforn , carbon-tetrachloride , acetone , hexamethylene , ethyl and methyl alcohols , had no appreciable induced magnetic double refraction .
I therefore examined some of these for a change in the value of on crystallization .
The results will now be briefly summarised .
All the experiments were made with the apparatus designed for low temperature work and most of the substances were investigated three times , the method being exactly as described in Part I. Carbon chloride , C. At the fusion point passed through a minimum value , as in the case of benzene , and , on further cooling the crystals , the susceptibility appeared to be the same as that of the liquid .
An effect of the same nature has been observed by HONDA* with sulphur , being a minimum at the fusion point 'Ann .
der Phys vol. 32 , p. 108 , 1910 .
CONSTITUTION AND TEMPERATURE ON MAGNETIC SUSCEPTIBILITy .
Acetone , This substance behaved like carbon tetrachloride .
Acetic acid , .COOH , and Propionic ).COOH .
Five experiments made on these fatty acids showed that the change of on crystallization was less than 1 per cent. in each case .
Hexamethylene , Three experiments showed that no change of takes place at the fusio1l point Chloroform , The susceptibility of the crystals did not differ appreciably from that of the liquid .
A control experiment on nitrobenzene gave the abnormally large value 12 per cent. ( which agrees with the earlier experiments with this substance ) for ( 7 ) A RELATION BETWEEN THE MAGNETIC DOUBLE REFRACTION OF ORGANIC LIQUIDS AND THE CHANGE OF MAGNETIC SUSCEPTIBILITY DUE To CRYSTALLIZATION .
A Comparison of Experir Results .
The general facts which have been described with regard to the large value of during the crystallization of aromatic compounds and the relatively small value which has been obtained in the later experiments for substances of an aliphatic nature form an interesting parallel with those relating to the magnetic double refraction of aromatic and aliphatic liquids investigated by MM .
COTTON and MOUTON .
The parallelism is particularly striking with nitrobenzene and hexamethylene .
For the former the induced magnetic double refraction and the value of are abnormally large while with the latter both effects are inappreciable .
I propose to give a table showing the values of the percentage change of susceptibility on crystallization and the magnetic double refraction ( M ) , each referred to nitrobenzene as unit , for a number of substances investigated by COTTON 'Ann .
de Chim .
et de Phys ser. VIII .
, vol. 19 , p. 153 , 1910 ; and se .
VIII .
, vol. 20 , p. 194 , 1910 .
'Journ .
de Phys ser. , vol. 1 , p. 23 , 1910 .
VOL. CCXV.\mdash ; A. and constitution .
Acetophenone , Benzene , Benzophenone , Bromobenzene , \mdash ; Br Chlorobenzene , Nitrobenzene , Toluenc , Xylcne ( Ortho ) , Aniline , Benzoylchloride , Pyridine , Hexamethylene , Carbon tetrach ] oride , C. Acetone , Acetic acid , .COOH Propionic acid , COOH Chloroform , Water , -bromonaphthalene is also an exception .
The value of is per cent. , while CorroN and MOUTON find that it shows a magnetic birefringence equal to that of nitrobenzene .
It is interesting to note in passing that not only does the Kerr electric effect follow the same laws as the magnetic double action with regard to variation of field strength and wave-length , but also that it shows a rough parallelism with for individual substances .
CONSTITUTION AND TEMPERATURE ON MAGNETIC SUSCEPTIBILITY .
and MOUTON and myself .
From the table it is clear that there is some connection between these two effects .
In each case the benzene nucleus plays an important role .
According to COTTON and MOUTON all liquids whose molecules contain such nuclei have a magnetic double refraction which is easy to detect experimentally .
These liquids also show a large value of o11 crystallization .
If the nucleus be modified by substitution of a monovalent atom or group for one of the hydrogen atoms the magnetic double refraction may increase or diminish but always retains the same order of magnitude .
The parallelism in this respect is well represented by the benzene derivatives investigated in Part I. All these show a value of of the same order except nzene .
In this substance the substituent , which is so active in augmenting the induced double refraction of the liquid , also contributes an abnormally lar ge part to as would be expected .
Both effects appear to depend also upon the degree of unsaturation of the substance , and if by any process we can lessen the number of double linkages associated with the nuclear carbon atoms there is a sudden diminution both of double refraction and .
Hexamethylene is particularly interesting from this point of view , for here ] six double are destroyed by the additional six hydrogen atoms .
In their later experiments COTTON and MOUTON have found that most organic liquids and some inorganic ones show a feeble double refraction in the gest magnetic fields .
These would be accounted for if the molecules possess a slight degree of dissymlnetry , for then they would becolne partially orientated by the external field .
As regards the change with such substances on crystallization , we should expect that the large molecular field which comes into play would produce a polarization or distortion in any molecule characterized by the molecule 's own dissymmetry , while the distorting force would depend upon the degree of unsaturation of the molecule .
Although the extent of this distortion is small , yet the forces which produce it must be large , for eve with the most unsaturated and unsymmetrical molecules amounts to a few per cent. only when the molecular field is of the order gauss .
With more delicate means of detection and a more intense magnetic field , COTTON and MOUTON conclude that a double refraction would be shown by all substances .
But this is exceedingly small compared with the natural double refraction of the crystalline medium , as far as experimental data on this question are available up to the present , and this increase of double refraction in passing from the liquid to the crystalline state demands a value for the local molecular field large in comparison with the strongest magnetic field which we can produce in the laboratory ( S4 ) .
It is expected too , that if we could use a more delicate means of measuring we * A large number of ordinary and extraordinary refractive indices are given in ' Recueil des Constantes Physiques , ' published by the French Physical Society .
These , however , refer to minerals which occur in sufficiently large crystals for optical experiments to be readily carried out .
Among the few organic crystals which have been investigated are cane sngar , ; benzile , should be able to detect a change of magnetic property when every substance crystallizes .
This again implies an enormous value for the local molecular field , even if the change of produced is so small as to be only on the verge of detection by our present methods , as the first part of S4 shows .
Among organic compounds aromatic substances show both these effects to an exalted degree and it is from a study of these that the large molecular forcives in diamagnetic crystalline media were first recognized .
By the argument of the present section it appears only justifiable to extend these results to diamagnetic crystalline media in general .
Such an extension is warranted also by the values of the latent heat .
ON THE NATURE 0F THE MOLECULAR FIELD .
The large local molecular field which has been recognized in diamagnetic crystalline substances must have its origin in the individual atoms or molecules .
In a liquid the effects of these forces at any point within the medium average out to a small resultant effect only as already indicated .
When the substance crystallizes the molecules are fixed in definite positions with regard to one another , and the polarization or distortion produced in a molecule owing to the forces exerted by its neighbours is no longer an average value .
The modifications of the physical properties of substances at the fusion point readily fall into line with this view .
When the substance is vaporized we must still assume that the force quite close up to the molecules is very large and rapidly falls off on account of the compensating action of the other half of the same molecule .
If the pressure of the vapour were sufficiently increased so that the local fields overlap , then we should expect that any physical property possessed by the vapour would become modified .
Thus has advanced the view that the pressure shift of spectral lines may be due to the co-operation of intense magnetic fields located in neighbouring atoms .
The pressure shift can be accounted for if the intra-atomic field , interpreted magnetically , has an intensity of the order gauss .
Further , RITZ has shown how to deduce the expressions of BALMER and RyDBEIiG for the representation of spectral series , providing we assume that the electrons are vibrating under an electromagnetic field whose order of intensity is gauss .
For this intensity of the intra-atomic field the frequency of vibration of the electrons corresponds to that of visible light .
If in the crystalline structure the surrounding molecules exert a local field of the order gauss , the electron in the particular atom we are considering would have its orbit modified by an amount which would account for a few per cent. change in the diamagnetic susceptibility .
It is not improbable that the intense intra-atomic fields of RITZ may be identified with the field due to the magneton , the molecular field being the result of their mutual action 'Astrophysical Journal , ' vol. 23 , p. 232 , 1906 , and vol. 35 , p. 268 , 1912 .
'Ann .
der Phys vol. 25 , p. 660 , 1908 .
CONSTITUTION AND TEMPERATURE ON MAGNETIC SUSCEPTIBILITY .
between the molecules of the crystalline structure .
In addition it is important f.rom our point of view to notice that these large fields are required by RITZ and HUMPHREYS in order to explain certain spectral phenomena , and therefore they are required apart from the magnetic nature of the atoms or molecules under consideration .
us now compare the theory which has been developed above and in Parts I. and II .
with the theory of ferro-magnetism given by WEISS .
The intense magnetic properties of iron below the critical temperature are attributed to the mutual effects exerted by the molecules of iron .
WEISS has shown that if we regard these effects as purely magnetic , then the large increase of specific susceptibility due to the transformation from the paramagnetic to the ferro-magnetic state may be represented by a molecular field , where is a constant of the order and is the intensity of magnetization for a field strength H. For saturated iron I has the value gauss ( approximately ) .
This internal field is called into action when an external field is applied , and its presence determines the abnormally large susceptibility of iron below the critical temperature .
In a similar way it has.been shown that the change of cific diamagnetic susceptibility when a diamagnetic substance crystallizes can be represented by a mean magnetic molecular field which is proportional to the intensity of magnetization ( diamagnetic moment per unit volume ) and whose constant of proportionality is comparable with N. On account of the zero moment possessed by a diamagnetic molecule initially , this mean molecular field is small relative to , but reasons have been given in previous work which show that the local value of this field is comparable with the ferro-magnetic molecular field .
It is this local field which distorts the electron orbits on crystallization and gives rise to a small change of diamagnetic susceptibility .
This and other considerations have shown that the intensity of the molecular field in diamagnetic substances is of the order gauss locally .
Up to the present the representation of this field as a magnetic field has not been justified except in so far as we have seen that it is in harmony with the important work of RITZ , whose molecular magnets have moments not inconsistent with that associated with the magneton , and whose magnetic field intensity is comparable with the ferro-magnetic molecular field of WEISS and the molecular field within diamagnetic crystalline substances .
It is important for our purpose to observe that the intra-atomic fields of RITZ are required to represent the distribution of lines in spectral series , whether the substance be ferro- , para- or diamagnetic .
Now very good evidence that molecular magnets do exist in ferro-magnetic has been obtained .
The fields due to these must contribute to the molecular field Taking the moment of the magneton as , we find for [ Added November 12 , 1914 .
] P. ZEEMAN , 'Researches in Magneto-optics , ' p. 178 .
the.saturation intensity of magnetization of iron which is nearly equal to the experimental value .
Therefore the increase of on passing from the paramagnetic to the ferro-magnetic state can be represented by a purely magnetic molecular field due to the mag1letons contained in the iron molecules .
This increase of is represented on WEISS'S theory by the molecular field gauss , and therefore gauss is the magnetic field due to the magnetons when the iron is saturated .
The continuity of magnetic induction demands that this field is to a large extent localized ( see p. 89 supra ) .
Now reverting to diamagnetic substances we are led , in the light of theory , to identify the local molecular field ( gauss ) in their case as due to the existence of molecular magnets , so arranged that for any particular molecule the moment is zero .
It seems that we could readily account on such a view for the fact that the chemical combination of two diamagnetic substances can give rise to a paramagnetic substance possessing magnetons .
This is the case for instance with the union of copper ( diamagnetic ) and sulphuric acid ( diamagnetic ) resulting in the formation of cupric sulphate ( which is paramagnetic and contains 10 magnetons to the molecule ) .
The forces which come into play during the chemical combination we may regard as upsetting the magnetic equilibrium of each component .
If these molecular magnets do exist in diamagnetic substances then the local molecular field of ths diamagnetic substance will be comparable with the molecular field in ferro-magnetic substances , and will be represented as far as they are concerned by a true magnetic field of intensity gauss ( approximately ) .
In this case the localized nature of the field is a necessary consequence of zero moment which has been assigned to the diamagnetic molecule .
In both ferro-magnetic and diamagnetic cases we have represented the molecular fields as due to gnetic force .
Now electrostatic forces , or indeed forces of any other nature , may distort the configuration of a molecule in the crystalline structure , and their effect could be represented by a magnetic field for both diamagnetic and ferro-magnetic substances .
Hence part of the molecular field for diamagnetic and ferro-magnetic substances may be a true magnetic field and part a magnetic representation of the distortion produced by forces of a different nature .
At any rate we cannot deny that the parallelism ich has been found between ferromagnetic and diamagnetic phenomena , taken in conjunction with the work of WEISS on the magneton and that of RITZ on spectral series , points to the conclusion that the molecular field in diamagnetics and ferro-magnetics is represented in part ( at least ) by a true magnetic field .
In this connection the introductory remarks of * An atom of iron has 11 magnetons , and the gramme-atom of iron contains 6 atoms , therefore 1 .
of iron ( whose density is taken as 8 ) wiJl contain atoms .
CONSTITUTION AND TEMPERATURE ON MAGNETIC SUSCEPTIBILITy .
Prof. J. W. NICHOLSON during the discussion on " " The Structure of the Atom\ldquo ; ( issued with ' Roy .
Soc. Proc. ' for July 20 ) are interesting : ' ' I think that Prof. RUTHERFORD has made it clear that the nuclear atom is , as a matter of fact , the only basis upon which profitable discussion of the constitution of the atom can really be carried on .
The main rival theory , as Prof. HICKS has just indicated , is the magnetou theory .
It seems that just recently in the course of a discussion it has been shown that the magneton theory and the nuclear theory do probably alnount to exactly the same thin Future work will test this .
But in the meantime it is clear that the magnetic forces cannot be neglected .
] Note February 1 , 1915 .
While this work has been passing through the press , some further extensions have bsen made .
In particular , there is one which I should like to refer to here because it bears directly upon the foregoing work .
It can be shown that the magnitude of the change of volume observed on crystallization may be interpreted as the -striction effect of the local molecular field .
We have seen that the natural double refraction of a crystalline medium may be represented as due to the complete orientation of the molecules by the local molecular field .
As the magneto-striction effect produced in a liquid and the induced magnetic double refraction are each proportional to the square of the applied field intensity , these results mutually support one another .
The further extension of the work is concerned with the validity of the quantum theory as applied to magnetic phenomena .
A discussion of the nature of the magnetic properties of HEUSLER'S alloys , from the points of view developed in the present work , will also be given .
I hope to publish an account of these extensions in a future communication .

