]\gt ; .
TOWNSBND , JOHN S. Magnetization oPPhiL Trans A , vol. 187 , pp. 533-649 .
Induced currents in a cylindrioal body and their effect on a coil round the centre , mathematical determination of .
Townsend , .Tohn S. Phil. Trans. A 1896 , vol. 187 , pp. 533-549 .
S$ Induction balance , a sensitive method of arranging an .
, John S. Phil. Trans. A 1896 , vol. 187 , pp. 5 Magnetism induced on surfaoe of a cylindrical body , mathematical investigation of .
Townsend , John S. Phil. Trans. A 1896 , vol. 187 , pp. 536-543 .
Magnetization of liquids , experimental determination of coefficient of .
Townsend , John S. Phil. Trans. A I896 , vol. 187 , pp. 533-549 .
Magnetization of liquids , effect of temperature on coefficient of .
Townsend , John S. Phil. Trans. A 1896 , vol. 187 , pp. 647-649 .
: [ 533 ] XIV .
of Liquids .
JOHN S. TOWNSEND , .
( Dubl .
) , vendish L , Cambridge .
Communicated by Professor J. J. THOMSON , F.R. Received June ll , \mdash ; Read June 18 , 1896 .
THE magnetization of liquids is a subject which has been examined by several methods , and although the relative values of the coefficient of magnetization for different bodies agree tolerably accurately , still there are large discrepancies in the absolute values given by different observers .
An account of nearly all these methods is given in WIEDEMANN'S ' Electricitat , ' vol. 3 .
The difficulty of finding the magnetizing force is no doubt a serious objection to the accurate determination of the coefficient in absolute measure by magnetometer methods , which , on the other hand , are very convenient for finding the relative values of for different bodies when magnetized by large forces .
The following null method of arranging an induction balance to determine in absolute measure gives very accurate results , and is very sensitive even when the forces are as small as 1 C.G.S. unit .
The apparatus which was used consisted of a primary and a secondary reuit .
quantity of electricity is induced in the secondary when the current in the primary is changed from to being the mutual induction of the two circuits , and the resistance of the secondary .
A commutator is used to commute the con nection of the primary with the battery , and at the same time to alter the connections of the secondary with the galvanometer in such a way as to send the induced currents in the same direction through the galvanometer , so that it will give a deflection corresponding to a current being the number of times per second the direction of is changed .
By this means a very small value of can be detected .
The mutual induction consisted of three parts:\mdash ; 1 .
The induction of a long ( wound on a glass tube ) on a secondary wound round its centre .
2 .
The mutual induction of two large circular coils . .
Au adjustable induction which could be altered by small known amounts .
21 .
10.96 Let internal area of the section of the solenoid tube .
number of turns in the primary coil .
number of turns , in the secondary coil .
length of primary .
The mutual induction is changed to by introducing a liquid into the ] enoid whose coefficient of magnetization is is no longer zero , as the deflection of the galvanometer indicates , and if is the amount by which must be diminished so as to reduce this deflection to zero we have The balance being restored .
We have thus a means of measuring in terms of quantities which are easily found without involving the formula the rate of rotation of the commutator or the value of the primary current .
The method is therefore very easily adapted to determine whether varies with the force , since the latter can be changed by alteriug the resistance of the primary circuit .
The general plan of the arrangements is shown in fig. 1 .
The continuous line denoting the primary and the dotted line the secondary circuit .
The solenoid A is connected with the other parts of the apparatus as shown .
The two circular coils are separated by an ebonite sheet ; the upper one belonging to the secondary circuit could be moved about till the mutual induction of the pair cancelled the mutual induction of the two coils A. The adjustable induction is represented at ( and also in fig. 2 ) , the primary circuit leading to the merculy cups at by which the required changes in induction are made .
and represent the commutator and galvanometer .
The sensitiveness is not much diminished by increasing the resistance of the primary , as it will be shown that in this case the rate at which the commutator revolves can also be increased without any danger of the Foucault current generated in the liquid having any effect on the secondary current , as as its effect on the galvanometer is concerned .
It is convenient to have a key , , in the primary circuit by means of which the current can be sent in either direction to the terminals , , so as to reverse the MR. J. S. TOWNSEND ON MAGNETIZATION OF LIQUIDS .
direction of the current in the secondary circuit .
'Fhe deflection of the galvanometer could thus have been increased by the multiplication method , but in all the experiments which were done the apparatus was found sensitive enough to give correct to the seventh decimal place by the ordinary deflection .
Fig. 1 .
Fig. 2 .
The great difficulty in the experiment is to ensure the perfect insulation of the secondary circuit , since the great electromotive force of the primary at break will cause a current to leak to the secondary if good non-conductors are not used to .
separate the two circuits .
It is necessary that , and should be at long distances apart , so as to satisfy the two following conditions:\mdash ; 1 .
The primary circuit of must have no lines of force through secondary of A. 2 .
Changing the number of coils in the primary of should have no effect on the galvanometer when is cut out of the secondary circuit ( by joining and ) .
Details of The primary circuit of the solenoid contained 1509 turnoe , wound in four layers on a cylindrical glass jar 47 centims. long and about 4 centims. internal radius .
The of 841 turns , was carefully insulated from it by means of several layers of paraffin paper , and occupied a length 4 centims. each side of the centre .
The mean internal sectional area at the centre perpendicular to the axis of the tube was found by rving with a cathetometer the rise in level of water in the tube , when a known additional volume was added to it .
The area thus obtained was sq .
centims. Let radius and half the length of the solenoid AB .
The force due to unit current is less than by the amount contributed by the two productions AA ' and BB ' of the solenoid , and being theoretically at infinity .
The solenoids , and , can be replaced by uniform surface distributions of magn etism , and , over the plane circular ends at A and , as far as forces at places external to and BB ' are concerned , being the number of turns per centim. * Hence the force at the centre of the solenoid is less by that due to a pole , at A and at , so that .
The second correction for the ends of the solenoid is found by considering the induction through the secondary circuit due to the magnetism induced on the Qurface of the liquid .
In cases where is small this correction is easily made since terms rolving k may be neglected .
* MAXWELL , ' Electricity and Magnetism , ' , sect. 676 .
MR. J. S. TOWNSEND ON MAGNETIZATION OF LIQUIDS .
537 Let be the potential of the magnetic force ; the element of the normal to the boundary of the liquid .
There is a distribution of magnetism on the surface of the liquid , where , which gives rise to a small force , at , in the direction opposite to F. So that , if is equal to the mean area of the secondary cOil and its number of turns , there is a diminution in the total flux of induction through the secondary.circuit equal to , due to the magn etism induced on the surface of the liquid .
Since , and invoJves as a factor , we may put , and the correction reduces to Now , since , we may , in determining the normal component of the magnetic force , omit and other small corrections and consider the case of a long solenoid of length and find the force at the ends parallel to axis , and the normal force at the cylindrical boundary , when the space inside is occupied by air .
It is easy to see that being measured from A to along the axis , and the potential of the distribution at , and at , so that at the ends of the solenoid the component of the magnetic force is at A or B. The foJlowing simple consideration also shows that the force parallel to the axis , at the ends is approximately half its value at the centre .
The force at due to the long solenoid AB ' is .
This must be contributed to equally by the parts AB and BB ' , hence , when the latter is removed , the force parallel to the axis is The double integral vanishes over any closed surface within which at every point , so that by considering the closed surface made up of the two planes perpendicular to the axis at and , and the intermediate part of a cylinder of section whose generators are parallel to the axis , and integrate over it .
We get over the plane end at over the ) of the cylindrical surface and the plane end at B. But at , and at B. Hence over the length of the cylinder , where area of the section of the cylinder .
Now let cylinder be a body whose coefficient of magnetization is small , and let it extend from A to , since the surface distribution is , we see that : When a current flows through a solenoid , and magnetizes a cylindrical body of equal length whose value for is small , there are uniform distributions , and - , induced on the plane ends , and of the total quantities , and - , induced , one-half resides on the ends and the other half on the cylindrical surface .
This latter distribution could not be got rid of by producing the ends of the cylinder beyond the magnetizing coil .
The actual value of the normal force at a point distance from the end of a solenoid of circular section of radius is given by the equation and being the complete elliptic integrals of the first and second kind , the modulus being may also be found by differentiating the expression for V in THOMSON and TAIT'S ' Natural Philosophy , ' .
If .
, Section 546 ) .
Making all the corrections , it is found that the change in induction is Substituting the following values , ; ; ; ; ; we get , therefore The adjustable inductance was made in two separate parts , and , shown in fi .
The larger changes in induction were made by , the secondary circuit of which consisted of seven turns wound round the inner bobbin , centima in diameter , and had its terminals at On the outer surface of the wooden ring , 37.17 centims. in diameter , were wound 10 turns of wire of small diameter close beside one another .
The ends of 1 , 3 , and 6 of these circles of wire were twisted together and terminated in 6 mercury cups .
The ripg and the bobbin were placed in concentric and coplanar positions , and were then fixed rigidly together by means of ebonite bars . .
J. a TOWNSEND ON MAGNETIZATION OF LIQTTIDS .
The second part of was similar in principle , and by means of it smaller changes of induction ] be made .
It had only 1 turn in the secondary , the 6 ends of 1 , .3 , and 6 turns on the outer ring terminating in 6 mercury beside those belonging to the larger inductance .
The radius of the inner circle of wire was , and that of the outer circles 25 centims. The planes of and were perpendicular to one other , each going through the centre of the other , so that when the number of coils of in series with the primary was altered there should be no change in the magnetic force perpendicular to the plane of , and vice versed .
The induction of one circle of the primary on the correspondin secondary is , for , 707 , and for urrents .
Objections have often been made to the induction method of finding by using a commutator , as it was supposed that , in the case of conductors , the induced currents had an ppreciable effect on the current through the secondary circuit .
This , however , is not the as when a galvanometer is used in the secondary circuit , but would introduce an error if a telephone were substituted .
order to prove this , let us consider the case of currents generated in a circular cylinder by the current starting in the primary , , and being the E.M.F. resistance and self induction of the primary .
We see ffom symmetry that the induced currents are in circles round the axis , and , if is the perpendicular distance of any point from the axis , and the strength of the current per unit section , ( rt ) .
Considering the circuit of radius , and thickness , extending through unit length of the cylinder , the current flowing round the circuit is , the resistance of the circuit is , where is the specific resistance of the liquid .
Therefore where is the totaI number of lines of induction going through the circuit . .
Hence we have *MAXWELL , ' Electricity and Magnetism , ' vol. 2 , chap .
iv .
, App. MR. J. S. TOWNSEND ON MAGNET[ZATION OF LIQUIDS .
( 1 .
) Differentiating with respect to , and by and differentiating a second time , we get Let , where is a function of only and as yet undetermined .
ITence we have : Hence where is BESSEL'S internal function of 1st order .
Hence substituting in ( 1 ) and using the equations and we get Hence , except when , in which c\amp ; se , The other coefficients are determined by the condition Chat , when From what follows , however , it will be seen that it is unnecessary to determine them , they do not appear in the result .
The number of lines of through the secondary due to the induced currents the liquid is *FORSYTH'S ' Differential Equations .
' MR. J. S. TOWNSEND ON MAGNETIZATION OF LIQUIDS .
541 The equation determining the secondary current is where consists of a number of terms of the form The solution of is , where and the quantity is Let us consider the coml)ined effect of the increase in permeability , and the currents generated in the liquid , on the quantity The first gives rise to the term , and the second to the series in N. Hence Substituting and bearing in mind that , we get ( 2 ) In order to select the important terms from this equation we must find the approximate values of , and The smaller values of the roots of are 2.404 , 5.520 , 8.654 , ' that the smallest value of is given by the equation *Lord Rayleigh , ' Theory of Sound , ' Vol. 1 , set tion 206 .
therefore and taking , we get so that , and about the same value .
Hence we see that in equation ( 2 ) , if we seek the values of for times greater than of a second after the primary current has been made , it is evident that te.rms having as a multiplier may be neglected in comparison with those multiby and , so that the terms in the sum 2 , reduce to Heuce : \mdash ; ( 3 ) ' and Hence the quantity induced in the secondary due to the presence in the solenoid of a liquid having conductivity is ME ' and the ratio of this to the quantity induced owing to a value of is , is increased when the resistance of the primary is increased , its smallest value bein , which is also about the value of MR. J. S. TOWNSEND ON MAGNETIZATION OF LIQUIDS .
543 , and when is small , so that the above ratio reduces to : and letting , which is about the time the commutator allowed the secondary to be connected in the positive direction , say , with the galvanometer after the process of changing the current in the primary from to was completed , we get a value for this ratio far less than unity in fact , the commutator might have been going twice as fast without the currents induced in the liquid having any effect on the value of in the eighth decimal place .
Results of the Exf'eriments .
In carrying out an experiment for the determination of , it is necessary in cases to syphon the liquid out of he solenoid immediately after been found , and test the balance again to see whether it is restored by replacing .
When this condition is fulfilled , it is evident that the mutual induction of the system of coils has not been altered while the experiment has been going on .
It does not always happen that an exact balance can be got by the method of changing the number of turns in the primary of , for if a certain adjustment gives a deflection of two divisions positive , a reduction of one turn in the primary of gives a deflection of three divisions negative , so that in this case the correct reduction can be taken as two-fifths of 65 , which is sufficiently accurate , since ] that is necessary in order to determine in the 7th decimal place is to be able to detect an induction of 20 C.G.S. units .
In all the following experiments the magnetising force was varied from 1 to 9 C.G.S. units , and in no case was there any change in Chloride .
The values of were found for four solutions of , the quantities of the salt per cub. cemtim .
being proportional to 1 , 2 , , and 4 .
The values of were 69 , 148 , 220 , and 298 .
These numbers are not proportional to , 3 , and 4 ; but when we add ( the value for the solvent ) to each , we get values for which are proportional to the weights of iron per centim. This shows the action of the solvent is to diminish the value of by , which is the value for water The actual values distiIIed water in three different experiments were , and the mean being The values for for water given by different observers HENRICHSEN QotNCKE - HoWARD - , and WABNmR A solution of ferric chloride , whose value for at centigrade , .
was 230.5\mdash ; 7.7 , was found by gravimetric analysis to contain0865 gram of iron per cub. centim. Hence the general formula for , for ferric chloride dissolved in water , is being the weight of iron per cub. centim. of the solution .
Ferric Sulphate .
The values of for three different solutions of ferric sulphate were found , the quantity of iron per cub. centim. in each being proportional to 2 , 3 , and 4 , giving the values 52 , , and 110 , for If to these we add we get the values of for the salt alone which are in the same proportion as the quantity of salt per cub. centim. A solution , whose value for at centigrade was , contained 0320 gram of iron per cub. centim. , which gives for ferric sulphate the formula It will be seen from the curves showing the variation of with temperature that at centigrade the value of will be per cent. less than at , so that we get , at centigrade , exactly the same formula as for ferric chloride .
* WIBDENANN , ' Eleotricitat , ' vol. 3 , pp. 1292 and 1822 .
: MR. J. .
S. TOWNSEND ON MAGNETIZATION OF LIQUIDS .
Ferric Nitrate .
An experiment was tried with ferric nitrate and estimated gravimetrically , like the chloride and the sulphate , and it also gave the formula thus showing that the coefficient depends only on the quantity of iron present , and the nature of the solvent .
Ferrous Chloride .
A solution of , which contained0409 gram of iron per cub. centim. , had a coefficient of magnetization given by the formula at centigrade .
The solution had been boiled for several hours with iron wire , so as to reduce any ferric salt to the ferrous state and then filtered .
So that for ferrous chloride we have the formula Ferrous Sulphate .
A more complete examination was made of fel.rous sulphate , the values for for two solutions , one containing twice the amount of iron per cub. centim. as the other , were found to be , and so that when they are corrected for the water in which they were dissolved they have the ratio 2 : 1 .
A solution was examined similarly to the ferrous chloride and gave a formula , the quantity of iron having been estimated gravimetrically by boiling a specimen with nitric acid and proceeding in the ordinary way .
this formula did not coincide with that got for the ferrous chloride a more complete examination was made of a solution which contained a small amount of ferric sulphate .
Having found the value of , 25 cub. centims. of the solution were extracted and the quantity of ferrous salt found by a standard solution of permanganate of potash , another specimen having been treated with nitric acid gave the total quantity of iron present , thus providing a means of finding the amount of ferric iron in the solution .
The numbers thus obtained were : : Weight of iron in ferrous state , per cub. centim. , 0689 , Weight of iron in ferric state , per cub. centim. , 0064 , subtracting 17 , which is the value of for 0064 gram per cub. centim. in ferric we get for 0689 of iron in the ferrous state .
Hence differing from the formula obtained for ferrous chloride by only 5 per cent. The same result was obtained for a solution containing01 gram of ferric iron and 0118 of ferrous iron .
The presence of free sulphuric acid in the sulphates was found not to affect the value of , by testing whether an addition of the acid altered it .
The liquid containing the two sulphates in nearly equal proportions was further examined by gradually neutralizing the free sulphuric acid with caustic potash till an almost neufxal solution was obtained .
This did not alter the coefficient of magnetization of the salt , showing that no change in the acid radical had as yet taken place , but when the potash was added till the dark precipitate was formed , the value of immediately rose .
This experiment shows that a salt corresponding to the magnetic oxide does not exist in solution , and that when the latter is dissolved in an acid it gives a mixture of ferrous and ferric salts .
Salts in the Dry State .
Estimations were also made in the dry state by filling a glass tube with the salt and putting it into the solenoid .
These experiments could not with ease be perfornled with the same accuracy as those made on the solutions , since the density of the salt is variable along the tube , and the amount of moisture and water of crystal lization in the body introduce errors when estimating the weight of iron .
The experiments showed roughly that the magnetic properties of the salts are the same as when they are in solution , the actual numbers got for ferrous sulphate were , the amount of iron per cub. centim. , found by weighing the body as , was 241 .
This value of differs by only 4 per cent. ffom the value calculated from the formula W. in the Acid Solutions of potassium ferricyanide and ferrocyanide each containing 029 gram of MR. J. S. TOWNSEND ON MAGNETIZATION OF LIQUIDS .
iron per cub. centim. gave values of the same as if water alone were present , so that when iron appears in the acid radical it is at least 100 times less magnetic than when it takes the part of a metal in a salt .
Effect of Temperature .
In estimating the effect of temperature it is necessary to use a tube of about onethird the sectional area of that on which the solenoid is wound , so as to be able to prevent the latter from being heated by radiation and vection .
The time required to keep the hot tube inside the solenoid while adjusting is about half-a-minute , so that it is an easy matter to prevent the coils tting heated .
The efficiency of the method employed was proved by the balance returning to its original zero when the liquid was taken out and the reduction in the turns in restored .
Uniformity of temperature along the column of the liquid is secured by inverting the tube containing it before inserting it into the solenoid .
The results of these experiments are best shown by drawing curves , fig. 3 , the ordinates denoting temperature centigrade , the ordinates being proportional to length corresponding to 50 being acted from the latter for convenience of representation .
The first curve was got from done on a solution of ferric chloride containing .
of iron per cub. centim. , the 2nd from felTOlS chloride containing .
of iron per cub. centim. , the 3rd ffom ferric containing amounting tegree centigrade aower teeth curves sthere ireat d chloride contalnlng 1 .
of iron per cub. centim. The curves show that , where is a function of the temperature , and is very approximately independent of the acid radical .
The diminution in density of the salt due to expansion contributes such a small amount to the rapid fall in that it is not necessary to make any correction for it .
The solvent itself did not show any variation due to temperature which could be detected by this method , and as the alcoholic solution gives the same temperature coefficient as the solutions in water , it is probable that the change due to temperature is a property of the molecules of the salt itself The first conclusion that the above results leads us to is that the magnetization is entirely due to the and is accurately the same for all acid radicals connected with it , so long as the iron remains either in the ferrous or the ferric state .
The only other electromag.netic phenomenon we know of that also possesses this property is the atomic charge , so that possibly the polarity of the metallic atom may be due to its rapid rotation carrying with it its atomic charge .
The ratio of the atomic charge in the ferric to that in the ferrous state is 3 : 2 .
There is a variety of suppositions as to the nature of the rotations of the iron atom in the molecule which will account for the ratio 266 : 206 of the two values of A simple case presents itself by supposing the intensity of the polarity proportional to and in the two cases , and denoting the charges , and and the rotations , the axes of which coincide with the axis of polarity .
In this case , if is the applied force , we get the displacements proportional to and , so that the ratio of the values of is : 206 , so that The next consideration is the controlling force , which acts in such a way that the induced magnetizat , ion is proportional to the applied force .
It cannot arise from any action of the solvent , since the magnetic properties of the salts are the same in the dry state as when they are in solution .
Also , since is accurately proportional to the density , we see that each lecular magnet behaves in exactly the same way whether the neighbouring magnets are near it or not .
It is , therefore , highly improbable that the surrounding molecules contribute to the controlling force , as there is no variation in it when the mean distances undergo large chang We are , therefore , led to the conclusion that there is no controlling force due to external bodies , and that the magnetization .
is a phenomena due to the perturbations in the angle of inclination of the axis of polarity to the direction of force H. Let denote this angle , which is a function of the where is the value of when , and 89 the perturbation at any time due to , and is proporMR .
J. S. TOWNSEND ON MAGNETIZATION OF LIQUIDS .
549 tional to , so that SHf ( t ) where differs from the different atoms according to and the initial circumstances .
The intensity I of magnetization is equal to where is the magnetic moment of the molecular magnet , and a time which is large compared with the periods .
Therefore , HE This explanation accounts for diamagnetism , for if we consider the very simple case of a magnet rotating rapidly in the hori ontal field of the earth , the effect of the applied force will be to make the north pole move more slowly when it is south of the axis of suspension than when it is north , so that its mean position is south of the axis of suspension , thus causing the magnet to act like a diamagnetic body .
The following table shows the values of for sa.lts in solution , being the weight of salt per cub , centim. , the forces ranging from 1 to 9 C.G.S. units . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
749 , ,

