]\gt ; The Measurement of Magnetic Susceptibitities of Low Order .
By ERNEST WILSON , M. Inst. C.E. , M.I.E.E. , Siemens Professor of Electrical Engineering in the University of London .
( Communicated by Prof. J. W. Nicholson , F.R.S. Received September 3 , 1918 .
) 1 .
Introductory .
The magnetic susceptibilities of various kinds of rock specimens have recently assumed considerable importance in relation to the new netic survey of the British Isles , and in this connection as a member of the Iron Ore Committ , ee of the Conjoint Board of Scientific Studies , the author was asked to endeavour to obtain a solution of the problem , which , in its essentials , consisted mainly in the development of some instrument capable of measurement of susceptibilites of low order .
At the suggestion of the Committee the author has prepared the present communication which contains an account of the instrument used , together with some of the results to which it has led .
The particular work on rock specimens which was the origin of the enquiry is given in\ldquo ; A Report on Magnetic Disturbances in Northamptonshire and Leicestershire and their Relation to the Structure * The type of instrument which has been used in the work described in the present communication is a modification of that known as the Curie balance , and depends for its action upon the pull exerted by a magnet .
2 .
The Magnetic Circuit .
The core of the magnetic circuit is in the form of a ring cut from a plate of a 3 per cent. silicon alloy of iron known by the name of Stalloy .
This material was chosen because of its high permeability at low values of the magnetic force .
The ring has an outside diameter of 18 cm .
and an inside diameter of cm .
Its thickness is cm .
An air-space with parallel sides is cut in the ring and its width is cm .
The poles are tapered at an angle of with the plane of the ring , and the opposing faces are 1 cm .
wide .
The ring is wound up to the pole pieces with five layers of copper wire of diameter cm .
, insulated with two layers of cotton .
The numbers of turns on the layers from innermost to outermost are 184 , 182 , 176 , 170 , 161\mdash ; tota1873 .
'Phil .
Trans , vol. 219 ( 1919 ) , Appendix .
' Annals de Chemie et de Physique , ' , vol. 5 ( 1895 ) ; 'Journal de Physique , ' , vol. 2 ( 1903 ) ; 'Proc .
Phys. Soc London , vol. 22 , pt .
3 , pp. 343-366 .
VOL Prof. E. Wilson .
The Measurement of The calculated ratio of the force in the air-space without the inserted specimen to the magnetising current in amperes varies from 676 to 725 owing to the changing permeability of the stalloy as the force in the airspace varies from 8 to 1000 C.G.S. units .
This is in close agreement with experiments made with an exploring coil inserted in the air-space and a ballistic galvanometer .
3 .
Description of the Instrument Apparatus .
The electromagnet ( fig. 1 ) wlnich produces the magnetic field is supported in a vertical plane on a wooden base which is capable of motion to and fro FIG. 1 .
on brass rails , 1 , 1 , of circular section .
The movement is controlled by a screw , 2 , having 20 threads to the inch which is turned by hand .
The winding on the electromagnet is continued by flexible conductors to terminals , 3 , fixed on the base at the back of the instrument .
The specimen , 4 , whose susceptibility is required is supported in a grip made of thin sheet aluminium or tin , which forms the extremity of a beam , 5 , made of aluminium wire cm .
diameter .
The horizontal distance between the centre of the vertical portion of the beam .
and the centre of suspension is cm .
which is therefore the effective radius .
The horizontal portion of the beam passes through a piece of sheet aluminium , 6 , which is suspended from the headpiece , 7 , by a phosphor-bronze strip , 8 .
The strip is 10 cm .
long and of rectMagnetic Susceptibitities of Low Order .
angular section mm. mm.* The head-piece consists of two concentric tubes of brass , the outer one being fixed to the frame work , and the inner one capable of being raised or lowered by a ring , 9 , which has a screwed thread on the inside , and engages the thread cut on the outer surface of the inner tube .
Two screws , 10 , 10 , fixed the inner tube and working in vertical slots in the outer tube prevent the .
tube from rotating .
The specimen is counterbalanced by an aluminium sliding weight , 11 , and an aluminium rider , 12 .
The mirror , 13 , is fixed to the aluminiunn sheet which supports the horizontal beam , 5 , and the moving system is damped by submerging an extension of it in an oil bath , 14 .
In order that the strip , 8 , may be maintained in a vertical position , four sighting pins , 15 , so arranged that the intersection of the two straight lines joining obliquely opposite pins coincides with the true centre of suspension .
The instrument is closed in at front and back by sheet glass which is capable of being moved upwards or downwards after the manner of an ordinary chemical balance .
A storage cell battery is employed to supply the netising current which is measured on an ampere meter .
An adjustable resistance is in cluded in the circuit in order that the magnetising current can be conveniently adjusted to produce the required magnetising force in the air-space of the electromagnet .
A reversing switch is employed primarily for the purpose of the magnet by reversals when variations in the force are required , and for removing all traces of residual netism .
4 .
The ld of the Preliminary experiments were made before deciding upon the shape of the pole-pieces , and much was already known in this connection from an investigation by Nagaoka , who has examined the distribution of the magnetic field in the neighbourhood of differently shaped pole-pieces by aid of the variation of Zeeman effect at different points of the field .
Unfortunately this method ceases to be applicable when fields so low as 3000 .
units are employed on account of diminished sensibility .
M. P. Curie ( .
cit in his original apparatus , employs two equal cylinders of iron with their circular ends as pole-pieces .
The axes of these cylinders are in a horizontal plane , and inclined at an angle of about , the distance between the nearest points being equal to the diameter of either cylinder .
Each cylinder is wound with a magnetising coil , and the magnetic circuit is For greater sensitivity a phosphor-bronze strip of rectangular section , mm. mm. has been employed .
' SAgaku-Buturigakkwai Kizi , ' 2nd Ser. , vol. 8 , No. 3 , p. 73 .
Prof. E. Wilson .
The Measurement of completed by yoke-pieces screwed to the base-plate .
arding the axis of as the direction in the horizontal plane bisecting the angle of inclination of the axes of the cylinders , and the direction at right-angles to this as axis of , it was necessary to find the value of where is the magnetic intensity at right-angles to .
A circular search coil was placed in the axis with its plane normal to the magnetic field , and on moving it quickly from point to point in the direction a series of deflections of a ballistic galvanometer was observed , from which could be found .
On account of irregularities it was found preferable to employ a different method based on the fact that , a necessary property , of course , of any magnetic field arising from a nlagnetic potential .
The search coil was therefore placed with its plane parallel to the direction of the field and normal to the direction , the displacements being in the direction In the present experiments a new method for the exploration of the field in the air space has been devised , which depends not on the determination of , but on the direct measurement of .
It is therefore in some respects essentially different ; and it is of course evident that the modification of the Curie balance , which admits such a direct measurement , must possess considerable practical advantages .
The arrangement is shown in fig. 2 .
A fine stlaight wire of platinum , 4 cm .
long , was fixed vertically to the extremities of two parallel aluminium wires , , lying in horizontal planes 4 cm .
apart .
The aluminium wires were fixed to a rod of ebonite , , which hung vertically , and was attached by a vertical conductor , , to the phosphor bronze suspension , whose force of torsion it was required to measure .
Looking at the aluminium wires in a direction normal to the horizontal planes , the two extremities above mentioned centrally situated between the poles of the magnet , , and in the mid-plane of the magnet ring .
The wires from these points are straight for a distance of cm .
in a direction normal to the plane of the ring , the object being to confine the force to the vertical platinum wire .
They then curve round towards the centre of suspension of the instrument , where they are attached to the vertical ebonite support .
this manner the platinum wire , , is capable of movement in a direction at right angles to its length , and for small arcs at right to the plane of the magnet ring .
A small mirror , , was attached to the axis of this system , and in conjunction with a spot of light and scale was used to measure the angular displacements .
An electric current was passed through the vertical platinum wire , , in the following manner .
A storage cell had its positive pole connected through an ampere meter and an adjustable resistance to the head-piece of the instrument .
The current could then flow through the phosphor bronze Magnetic Susceptibilities of Order .
strip to the upper of the two horizontal aluminium wires .
It bhen passed this wire to the platinum wire , , and thence to the lower aluminium wire back to the centre of support .
It could then flow down a vertical conductor , , dipping into mercury , Hg , which was connected with the negative pole of the cell .
For the purpose of speedily bringing the moving system to FIO .
2 .
rest , a vane , , of platinum foil attached to the system was submerged in the mercury .
When it was desired to measure the variation of the strength of the magnetic field of the magnet in a direction Ox at right angles to the plane of the magnet ring , an electric current , , was caused to flow through the Prof. E. Wilson .
The Measurement of vertical wire , , and was maintained constant .
The magnet was excited by a current , , and the deflection was noted .
The current A was then reversed in sign and the deflection noted .
The current A was then Broken , and the magnet was moved forward turning the adjusting screw through one revolution .
This produced a movement of cm .
In this way a series of observations was made , which rendered it possible to determine graphically the variation of the force exerted by the magnet in terms of the distance Ox from the central point of the air-space at rightangles to the plane of the ring .
5 .
Force on the Suspended System .
If the intensity of the magnetic field is constant and equal to , then the mechanical force in dynes which , acting on the vertical conductor , produces 1 degree of twist is equal to constant ; where is the length of the conductor , the current in the conductor , and the observed deflection .
A number of determinations were made , and they indicate a mean value of dynes per degree of twist .
A further test was made by suspending a horizontal bar of known moment of inertia , and observing its time of oscillation .
The details of these observations are not included , as they follow obvious lines .
The value of the force in dynes for 1 degree of twist at this radius is for a large bar .
For a small bar the force is dynes .
The average value obtained from these determinations is dynes , which is in close agreement with the value found by the electrical method .
6 .
Field of the Condnctor .
A possible disturbing factor which requires examination is the magnetic effect of the cur1ent , , in the vertical conductor .
This current was varied for the same strength of the netic force , , in the air-gap .
Conversely for the same value of the ; lising force , , was varied .
The constancy of the results and their agreelnent with those of the mechanical tests indicate that when the vertical conductor is within the air-gap the disturbance caused by the current is under the conditions of the experiments inappreciable .
The effect of susceptibility of the platinum wire was investigated separately .
7 .
On the Relation between Maynetic Force of Excitation and Magnetic Susceptibility .
Some light is thrown upon certain effects which have been observed during the course of the present experiments by a study of the magnetic properties Magnetic Susceptibilities of Low Order .
of iron filings which were tested by the author by the ballistic galvanometer ring method before and after heating to 90 C. * The curves showing the relation found in this way between the magnetic induction , , and the netic force , , do not deviate greatly from straight lines for values of up to 150 C.G.S. units or more , and they indicate average values of the susceptibility and respectively .
In the present experiments on certain of the rock specimens in connection with the Geological Survey , the relation between the magnetic force , , and the susceptibility is very nearly a straight line .
For example , with the partially oxidised oolitic ore ( N.88 ) the following results were obtained , and they illustrate a very close approximation:\mdash ; In the case of igneous rocks of high susceptibility there are indications of a maximum susceptibility at the intermediate forces .
The following figures were obtained for example , with dolerite from Whitwick , Leicestershire ) , granite , Mount Sorrel Quarry , Leicestershire , and augitegranophyre ( syenite ) , from Groby Quarry , Charnwood Forest .
8 .
The Instrumental Constant .
It has been shown that when the susceptibility of the specimen ( K ) does not exceed C.G.S. units , the magnetio force , , in the air space is proportional very nearly to the current in the magnetising coil .
Let the current in amperes be denoted by A so that is also a function of the distance measured at right angles to the plane of the magnet ring from point midway between the pole pieces , and lying in the mean plane of the ring .
Let this function be denoted by .
We have then where depends neither on the amperes , nor the distance .
Then if is the volume of the body the mechanical force , , exerted the magnetic field upon the specimen is given by KV The Electrician , ' October 5 , 1900 .
Compare Rucker , ' Roy .
Soc. Proc vol. 48 , p. 505 ( 1900 ) .
Prof. E. Wilson .
The Measurement of Therefore , for any given distance , , and since the observed deflection is proportional to , we have proportional to , or .
It follows that for unit volume the absolute value of the susceptibility , , is equal to the observed deflection , , divided by the square of the current in amperes and multiplied by a constant peculiar to the instrument .
That is to say ; where will be called the instrumental constant .
The determination of this constant is naturally the fundamental problem for any instrument constructed after this pattern .
It has been performed in various ways , only a few of which are indicated the sections .
9 .
Node of Determination of the Constant with Specified Material .
Five determinations of the instrumental constant have been made in which the intensity of the magnetic force in the airspace was varied from 221 to 2210 C.G.S. units .
In any one experiment the average values of the observed deflections and have been plotted in terms of the distance Ox at right angles to the plane of the magnet ring , and , from the curve so obtained , the values of the magnetic force in arbitrary units have been taken .
Let , be consecutive values of the magnetic force due to the magnet along the line Ox , separated by intervals .
It is required to find the average value of over a distance along Ox , covered by the specimen when the maximum deflection is obtained .
The value of adopted in the experiments is cm .
, and , if the specimen has a width along Ox of 1 cm .
, we may take it that the area of the curve included between ordinates separated by 1 cm .
, which give a maximum value , will approximately determine the equilibrium position of the specimen .
Since at any point between the extreme positions , supposed close together , defined by the suffixes zero to unity , it is necessary to square the values of , .
take the difference of successive pairs of these , and divide by 2 in determining the value of between these pairs .
The values of , in arbitrary units have been taken , and the mean curves plotted from the average deflections , as already described .
It is found , by direct measurement of areas , for example , when .
is equal to 469 C.G.S. units , as in Experiment No. 2 , that the mean value of , etc. , or between and inclusive , is 332 ; and the Magnetic Susceptibilities of Low Order .
mean value between and inclusive is 330 .
The specimen will , when the deflection is a maximum , be nearly symmetrical with the ordinate at the point defined by Ox , A constant of the instrument will now be worked out , making use of the mean value 332 thus obtained .
The scale used in the experiments is divided into 1000 equal divisions , counting from the extreme left , and it is curved to the radius of the spot of light .
maximum deflection is obtained in the first place by moving the magnet in one direction from its mean or central position , and then by moving it in the opposite direction .
Thus the difference between the two readings gives double deflections , which are an ultimate measure of the susceptibility .
If is the volume-susceptibility in C.G.S. units , the observed deflection in double scale divisions , A the netising current in amperes , and the volume of the specimen in cubic centimetres , then the instrumental constant for small values of is equal to .
One double deflection corresponds to a single deflection from the mean position of one half-division , or 1/ 40 cm .
The scale is distant 77 cm .
, and therefore the angle of twist if allowance is made for the deflection being doubled by the reflection of light .
The mechanical force in dynes per gree of twist is , so the total mechanical force is in this instance .
The maximum value of the magnetic force in the gap is 670 C.G.S. units for a value of of 1 ampere .
Thus , and the instrumental constant A summary of the results obtained in the five determinations is given in the following Table , and the mean value of the instrumental constant is Prof. E. Wilson .
The Measurement 10 .
Effect of Finite Size of the Specimen .
The above value of the instrumental constant strictly requires a further connection as no account has been taken hitherto of the fact that the specimen has an approximately circular cross-section of finite size .
Assuming that the lines of force in the specimen are straight lines parallel to the plane of the magnet ring , the mechanical force on an element of volume is .
Take as the element a slice of breadth , width 2 , and height 1 cm .
, and the force on the element is The total force for the specimen is accordingly For a cylindrical specimen of radius Therefore ; where is the radius of cross-section .
This can be evaluated by quadrature in a en- number of equal parts , but after evaluating it analytically by this method , on the assumption that the graph of against was approximately parabolic , and again , by direct measurements from a completely graphical treatment , it was found that the latter method was much more satisfactory .
In fact , the curve is not sufficiently parabolic to render any analytical treatment of greater value , for it is unsymmetrical on the two sides of its vertex .
The actual method adopted was therefore the construction of a graph of the integrand in the preceding formula , and the evaluation of the integral between any limits by direct measurement of the area included between this graph and the corresponding ordinates .
11 .
Determination of Constant by Use of Iron Ores and Alloys .
( a ) 13 per cent. Alloy of and Iron.\mdash ; A ring of this alloy , given to the author many years ago by Sir B. A. Hadfield , F.R.S. , having an internal diameter of cm .
, and an external diameter of cm .
, and a depth of cm .
, was tested with the ballistic galvanometer .
Its permeability was found to be for forces up to about Magnetic Susceptibilities of Low Order .
200 C.G.S. units , and this agrees with the value usually quoted .
* A piece of the ring was then cut out , so as to provide a straight specimen , having a length about cm .
, and square cross-section 1 cm .
cm This specimen had some of the original surface of the ring left on one side , which was evidently much more netic than the remaining three sides .
Owing to preliminary tests with a horseshoe magnet , the specimen was netised transyersely along this particular surface .
It was tested in the balance with the original surface of the to the plane of the crnet , and whereas the value of was 205 with the specimen oriented one way , it was only 30 when it was turned round through .
If there were no residual netism these values of should be equal .
It was then demagnetised from a maximum force of 2040 C.G.S. units at frequency 54 , and the values of were and and were still very unequal .
The riginal surface of the ring is undoubtedly highly retentive .
On grinding off all traces of the original surface the susceptibility of this piece dropped to from the average value which it had in the first instance , and from the value in the original ring .
Specimen No. 2 was cut from the same ring and had a circular crosssection of diameter 1 cm .
It had patches of the original surface of the ring along its whole length on opposite sides of a diameter .
In its first position this diameter was at right angles to the plane of the magnet ring , and the measured susceptibility was .
On turning the specimen through the susceptibility fell to .
On turning it further until it was completely reversed ( through ) the susceptibility rose to .
The the patches of the original surface is clearly evident .
Specimens 1 and 2 were then maintained at a temperature of about 53 C. for 100 hours with a consequent increase in the susceptibility on return to atmospheric temperature .
The circular bar No. 2 was tested in the first instance just as it left the furnace , and its susceptibity was , taking an average of two determinations .
On the specimen with emery and demagnetising it , the average susceptibility fell to 0.267 , which shows directly the effect of a very thin layer of oxide .
Specimen No. 1 had also a high susceptibility , but as it was desired to obtain a susceptibility of lower , its temperature was raised to about 75 C. and it was allowed to cool in air .
It was then polished on emery and demagnetised ; the susceptibility was then found to be .
The specimen was then ground so as to have a circular cross-section , its volume being reduced from to .
After demagnetisation the susceptibility was , which is larger than the value found with the square section , partly no 'Phil .
Trans p. 462 ( 1885 ) ; The Electrician , ' November 9 , 19(K ) .
Prof. E. Wilson .
The Measurement of doubt on account of the test , being only made with the sides of the square parallel to the faces of the pole-pieces .
The instrumental constants given by these two specimens in their final state have the values 132 and A third specimen ( No. 3 ) , having a circular cross-section , was cut from the ring , and not so much of it ground away as in the case of specimen No. 1 .
Its susceptibility had an average value of , which is larger than that of the No. 1 specimen before heat treatment , possibly on account of the presence of the original surface of the .
Its susceptibility was more than twice that of the No. 2 specimen when first tested .
Its density was appreciably smaller , probably on account of the existence of small blow-holes , .
of which there was distinct evidence .
It would appear that the susceptibility of this steel as received from the foundry , when all traces of the original surface have been removed , is of the order , that is to say , its permeability is of the order as against , the value usually quoted ( loc. ( b ) Spiegel Eisen.\mdash ; A small piece of this aJloy which was supplied to the .
author by Prof. A. K. Huntington , was ground so as to form a fairly regular test-piece with two of the crystal faces along its length .
The piece was cm .
long , and its cross-section measured cm .
between the faces , and cm .
at right angles thereto , the sides being approximately round between the crystal faces .
It was highly magnetic and very retenGive when received .
It was then heated to bright redness and allowed to cool freely in the air , in order to remove all traces of magnetisation .
Two determinations were made in the balance ( a ) when the smaller width was parallel to the plane of the netic ring , giving ( b ) when turned through , the deflection then being 247 .
figures obtained were as follows:\mdash ; The value of the susceptibility , as found in the magnetometer with was C.G.S. units .
So that the instrumental constant from this investigation was found to be : It is interesting to note that another piece from the same block ( not heated ) had a susceptibility of ; whereas another specimen ( also not gnetic Susceptibilities of Low Order .
heated ) from Prof. Herroun 's collection had a susceptibility of .
In both cases heating to redness for a few minutes reduced the susceptibility to about -third of its original value .
( e ) Slatey Ma.fanchuria ) piece obtained by permission from specimen 9500 in the collection of the Geological Survey and Museum , has a volume , and density .
When it was tested in the balance the following figures were obtained:\mdash ; gneticforce HDeIlection Across lVith l The mean value of is 1,100,000 , and the susceptibility , as iven by the magnetometer with , was .
The instrumental constant is herefore 12 .
The Use of Rock Specimens .
( a ) oiitic Iron Ore ( Dolgelly).\mdash ; This specimen is from the collection of the Geological Survey and Museum , and has a volume of .
and a density .
Its susceptibility , obtained by the netometer , is With a current of amperes , corresponding to a magnetic force equal to , the value of is 119,000 .
The instrumental constant becomes in this instance ( b ) lagnetite Calcite ( Aran).\mdash ; From the Geological Survey and Museum Collection .
The volume of this piece was .
and its density was It contained detached magnetite crystals .
Its susceptibility , as tested in the magnetometer , had the value .
An additional test was made by inserting the specimen in rectangular holes cut in the pole-pieces of an electromagnet having a core of stalloy .
An exploring coil was wound on the specimen and attached to a ballistic galvanometer .
Deflections were obtained on reversal of the magnetising current ( a ) with the specimen in position , and when the specimen was removed , and the exploring coil was supported in its original position on a wooden former having the same size as the test piece itself .
The ratio of the deflections indicated a susceptibility of When tested in the balance with the current , corresponding to a force , the value of was 1,370,000 .
The instrumental constant is , therefore , found to be or Prof. E. Wilson .
The surement of The work on this particular specimem is of interest in that a new method was adopted .
Hitherto the susceptibilities of the materials examined ere such that they could readily be tested independently by the direct method of the magnetometer .
In connection with the application of such tests , the w1iter wishes to acknowledge cordially the valuable assistance rendered to him by Prof. E. F. Herroun , F.I.C. , who devoted a considerable amoupt of time to this work and made valuable suggestions with regard to the materials suitable or available for the work .
In connection with the present specimen , the susceptibility was of such magnitude that the ring method with ballistic galyanometer became an experimental possibility , and therefore was adopted as an alternative to the magnetometer .
[ Note added November 13 , 1919.\mdash ; This method fully described , and further information on the susceptibilities of the Manchurian and other magnetites is given in a paper read before the Physical Society of London .
A portable instrument nsed in some of this work was shown at the meeting of the Royal Society , at which this paper was read .
For details see 'Proc .
Phys. Soc vol. 31 , Part 5 , August 15 , 1919 .
] ( c ) Dolerite from a Coal Shaft , Whitwick , Leicestershire 114 .\mdash ; Beport to the Survey and Museum , February 12 , 1918 .
The specimen has an approximately circular cross-section of diameter 1 cm .
, and its length is about 4 cm .
Its volume is .
and density .
In appearance this rock is much more uniform than , for example , the grey granite ( L15 ) from Mount Sorrel Quarry , Leicestershire , and its susceptibility is fairly high .
It appeared to be suitable for testing the constant of the instrument .
There are indications of a cleavage plane parallel to the axis of the cylindrical test piece , and on cutting a portion away a layer of very small crystals of netite was discovered .
It is apparent , therefore , that the susceptibility obtained is not representative of the whole mass of the specimen .
Demagnetisation was carried out in this case by the specimen in a solenoid and varying the force from a maximum of 1500 C.G.S. units to zero at a frequency of 50 .
The susceptibility was slightly increased , showing that there were residual effects in the specimen as received .
The average value of is 34,000 , and the magnetometer gave a susceptibility equal to .
The instrumental constant is therefore which cannot , however , be regarded as trustworthy .
13 .
The Use of Powders .
In a report to the Director of the Geological Survey and Museum , dated December 17 , 1917 , it was pointed out by the author that if rock specimens Magnetic Susceptibilities of Low Order .
are reduced to powder by a grinder , in the construction of which no magnetic material is used , and then tested for magnetic susceptibility by the instrument now described , the susceptibilities of the original specimens could be obtained .
The powders so prepared are placed in a glass tube of known internal volume , the closeness of packing being due to tapping the tube somewhat firmly at intervals during the filling process .
Two sizes of tubes have been employed , having internal diameters of cm .
and cm .
respectively , and the powder occupied a length of 4 cm .
in each case .
The superficial or apparent volumes of the contained powders are therefore .
and .
respectively .
To obtain the volume-susceptibility of the original rock it is necessary to the powder that was in the tube and to find the volume of the solid rock from whence it was obtained\mdash ; thus making use of the density of the solid specimen .
The permeabilities of powders have in some cases been determined by Crane .
* but his method , which is quite different , appears to be open to serious criticism , and it is not certain that the quantity measured is really the permeability of the specimen .
We do not , however , propose to compare the results obtained with those of the present paper , as it would not be readily possible , in view of the totally different of most of the materials used .
Accounts are now appended of some of the more typical powders dealt with , and their nlodes of treatment .
They may be regarded as illustrative cases .
( a ) Normal Grey Granite , from Jfouvnt Sorrel , icestershire ( piece of this rock ( not the specimen tested in the solid state ) was reduced to very fine powder by an iron grinder , and gave an extraordinarily value of the ceptibility which was about times the value obtained with a solid piece taken from the same hand specimen .
By chemical an alysis the powder was found to contain about psr cent. of metallic iron .
Another piece of the same rock was crround in a Wedgwood mortar , and this had a susceptibility which was times that of the original test piece .
Finally , the original test piece was finely ground in a Wedgewood mortar and its susceptibility when referred to the volume of the original rock was very nearly the same as was found with the specimen when solid ( see Table I ) .
These results show ( a ) that in the case of this granite the distribution of magnetite is not uniform , that is to say , pieces of the same rock may have different susceptibilities ; ( b ) if a piece of the granite be first tested in the solid state and then finely ground in a non-magnetic inder , the susceptibilities are found to agree fairly closely .
'Trans .
Amer .
Inst. Mining Engineers , ' vol. 31 , p. 405 .
Prof. E. Wilson .
The Measurement of ( b ) New Zealand Iron sand was given to the author some years ago by Mr. C. R. Palairet , and the iron produced therefrom was tested for its magnetic properties .
* The chemical percentage composition of the sand is as follows :Peroxide of iron , protoxide of iron , alumina , protoxide of manganese , magnesia , silica , titanic acid , oxide of cobalt trace , phosphoric acid none , sulphur none : total l A mixture of this sand , whose absolute density is , with hard paraffin whose density is containing 25 per cent. by volume of the sand , or 63 per cent. by weight , was used in the production of rods having diameters of and cm .
respectively .
The rods were 8 cm .
long and , tested in the magnetometer , gave values of the volume-susceptibility and in C.G.S. units .
The rods were then cut in half so as to form suitable specimens for the balance and the results obtained were as follows:\mdash ; When tested alone the volume-susceptibility of this sand is taken on superficial or apparent volume : has a mean value of 1,240,000 , and the instrumental constant is ( c ) Northampton Iron Ore \mdash ; This iron ore is very uniform and has yielded good results .
The method adopted was to first test a solid specimen , and then to grind it in a Wedgewood mortar and test it in the powdered state .
It will be seen from Table I that in the solid and powdered states the volume-susceptibility is almost exactly the same , namely , C.G.S. units .
14 .
On the Influence of Size of It will be noticed on referring to Table I that the susceptibility in the above cases is htly larger in the case of the smaller tube .
The theory underlying the action of the instrument indicates that this should be the case .
This point is tested further , and reference shouId again be made to the fures given in Table I. The red serpentine from the collection of Prof. E. F. Herroun shows a slightly larger susceptibility when tested in the smaller tube .
When referred to the volume of the original rock the ' The Electrician , ' September 17 , 1897 .
Magnetic Susceptibilities of Low Order .
Prof E. Wilson .
The Ieasurement of susceptibility of the powdered specimen slightly larger than when .
solid .
The dark green serpentine , also from the same collection as the red variety , shows a close agreement between the values for the solid and powdered states .
A specimen of red haematite was finely powdered and its susceptibility when referred to that of the original rock is slightly larger than in the solid .
To contrast with the figures obtained from the red haematite , has been tested .
If referred to a possible solid state the susceptibility of powder would be smaller than that of the haematite itself .
It is noticeable that the smaller tube gives the larger value of the susceptibility .
The effect of diameter in the specimen is very marked in the case of which has also been tested .
The only powder so far tested which gives a contradictory result is Fed .
solid tested in the magnetometer gave as the susceptibility .
The powder , when referred to the original solid , gives .
The instrumental constant 15 .
xperiments with Ftuids .
The fiuids with which experim have been conducted are solutions of anese sulphate and ferrous sulphate in water .
These were selected because a certain amount of experimetal information has already been obtained , which is of a somewhat contradictory nature , very diverse values for the mass susceptibility having been given by different observers .
The mass-susceptibility of a solution of such a salt in water is given by the formula Where is the -susceptibility of water , which may be taken as is the mass-susceptibility of the water-free salt , and the percentage weight of salt in the solution .
Typical experiments with the solutions led to the results :\mdash ; ( a ) Manganese Sulphate.\mdash ; The values given in the Landolt-Bornstein 1912 Tables for the mass-susceptibility of manganese sulphate are C.G.S. units at atmospheric temperatures .
The among these values is considerable , and serves as a valuable indication of the fact that susceptibilities of this order have not hitherto been determined with precision .
The mean of four values is C.G.S. units .
A solution of managanese sulphate containing .
of anese , or .
of sulphabe per cubic centimetre was placed in a test tube having Magnetic Susceptibilities of Low Order .
an internal diameter cm .
, the total volume of the solution being 2 .
After correction for the susceptibility of the tube itself , it was found that For this solution , , and using the mean value of , we obtain on calculation C.G.S. units .
The density of this solution is , so that the volume-susceptibility is with a corresponding value for the instrumental constant .
It is clear from the previous determinations that the constant is decidedly larger , suggesting that the value 85 for the nitude K is greatly in error , and that the mean of the others is more reliable .
Adopting these three only , the mean value becomes , from the work of previous investigators , whence which is very close to the value of the constant found from the instrument itself .
The third entry in Landolt-Bornstein thus appears the most reliable .
Ferrous Sulphate.\mdash ; Landolt and Bornstein 's Tables quote seven values for the mass-susceptibility of this salt : C.G.S. units , some of which must be very inaccurate .
We shall provisionally reject the last two in forming a mean , which becomes A solution used had density and a value of equal to .
We obtain and for the volume-susceptibility , The value of , as given by the instrument , and corrected for the glass , was l50 .
The resulting value of the instrumental constant is , but little reliance can be placed on it .
The solution contained .
of iron or .
of ferrous sulphate per cubic centimetre , the total volume being , as before .
The value of is probably about 93 , the highest previous estimate .
Prof. E. Wilson .
The Measurement of 16 .
The Final Instrumental Constant and Use of the Balanxe .
In the interpretation of the values of the instrumental constant various circumstances must be borne in mind .
For example , that given for manganese alloy of susceptibility can be of little value .
For the method is not really applicable to such large susceptibilities without considerably more mathematical development of the higher approximations .
The criteria , which , in fact , define the applicability of any direct method of measuring , are set out in Maxwell 's treatise , * and are well known .
When the susceptibility exceeds a certain small amount\mdash ; determined greatly by the diameter of the specimen\mdash ; the demagnetising forces must be accounted , and the simple expression for the mechanical force per unit volume cannot be used .
It was nevertheless of interest to include this value , as showing the extent of error involved .
The values for many rock moreover , are unreliable , on account of difficulty of securing an even approximately homogeneous specimen , especially as regards its magnetic content .
This source of error applies , to a greater or less extent , to all the specimens , but is likely to act in a fortuitous direction among the species .
The results with solutions have already been discussed sufficiently .
The general mean of the results is taken without the use of Spiegel Eisen and manganese alloy .
It is not , however , altered appreciably if they are included .
The average relates to rock specimens and minerals generally .
The general mean , taken as above , is , and no appreciable divergence occurs if the values are weighted according to any probable system .
The constant deduced from the instrument itself was and the general agreement may be regarded as satisfactory .
17 .
The Effect .
of Length and of Cross-section of the Specimen .
The variations in the form of cross-section and length of the specimens were investigated somewhat fully .
The results demonstrated that the dimensions selected were satisfactory , and it is not thought necessary to give an account of this part of the work .
18 .
Susceptibility of Varieties of Mica .
A representative series of micas was tested by the author some years ago for dielectric strength , capacity , and insulation resistance , and , as specimens of these were available , they have now been tested for susceptibility by the balance .
'Elec .
and Mag vol. 2 , p. 67 , Edition 1873 .
' The Electrician , ' December 16 , 1904 , and March 17 , Magnetic Susceptibilities of Low Order .
449 Prof. E. Wilson .
The Measurement of Each specimen was built up of laminae about cm .
long , and of varying width , so as to produce an approximately circular cross-section 1 cm .
in diameter .
There were from to 16 plates in each specimen , which } were held together by two turns of fine cotton round each end .
Each specimen has been tested with the direction of the magnetic field parallel to the planes of the laminar and right angles thereto .
On reference to Table II , it will be noticed that , in each specimen , the susceptibility is larger in the former case .
The ratio of the susceptibilities parallel to and at right angles to the plane of the plates is much more pronounced in the case of the spotted micas .
For example , Nos. 1 and 10 give the remarkable ratios and .
Spotted specimens contain oxide of iron as inclusions , and under the microscope distinct crystals of magnetite in laminar are visible , hence their high susceptibility .
Clear ruby mica .
( Nos. 14 and 17 ) is almost exclusively used for high grade condensers , and its susceptibility , though small , is not the smallest of the series .
The ruby ( Nos. 6 and 11 ) , much stained by kaolin inclusion , has very low susceptibility , as also has the yellow variety ( No. 13 ) .
The amber series ( Nos. 15 15 nave also a low susceptibility .
The results obtained in different directions would appear to imply an orderly arrangement of the included particles of oxide of iron , or an actual difference in behaviour of the mica itself in different directions with relation to the crystal axes .
The attempt has not been made in the present communication to decide between these alternatives , but the phenomenon is so marked that it appeared desirable to place it on record .
19 .
Susceptibility of Aluminiurn .
The physical properties of certain alloys of aluminium were examined in and since that time the specimens have been on the roof of King 's College , London , in order to investigate the effects of exposure .
Pieces of the original wires have been preserved , and a selection from these has been made for the purpose of investigations on susceptibility .
Each specimen , with the exception of No. 26 , has a length of 4 cm .
and a diameter of cm .
Referring to Table III , No. 25 is a piece of commercial aluminium , whose conductivity is per cent. of that of pure copper .
Using the constant the volume susceptibility is C.G.S. units .
The volume susceptibility of pure aluminium is stated to be and its mass susceptibility No. 21 was chosen on account of 'Journal of Proceedings of the Institution of Electrical Engineers , ' Part 164 , vol. 31 ( 1902 ) .
rnstein 1912 Tables .
Kaye and Laby Tables .
Magnetic Susceptibilities of Low its comparatively large iron content .
Comparison with No. 23 would indicate that manganese has the effect of increasing susceptibiliby .
It has been shown earlier that a normal solution of has a higher susceptibility than a similar solution of .
Again , it is well known that coppermanganese alloys of aluminium can have high susceptibility .
No. 22 has about the same percentage of nickel as No. 21 , but its iron content is much smaller .
This alloy ( No. 22 ) is remarkable on account of its high breaking load , which is 45 , .
per square inch of cross-section , as against 28,200 for commercial aluminium ( No. 25 ) .
There is also a great difference between the temperature coefficients of electrical resistance of these alloys .
That of No. 22 has the value as ainst 0 for No. 25 .
No. 20 has also a high breaking load square inch ) as compared with No. 25 , and its susceptibility is slightly higher .
The chief characteristic of specimen No. 26 , whose length is 4 cm .
, is the presence of per cent. of cobalt .
Its susceptibility is not much higher than that of pure aluminium .
The low value found in this specimen may be influenced by its relatively large diameter .
Table III.\mdash ; Alloys of Aluminium .
ityunits .
20 .
Note on the lities of Glasses in Relation to Composition .
In some recent confidential work on the electrical conductivity of a series of glasses as affected by temperature , the author had -an opportunity of using the balance as a detector of the amounts of oxides present in their compo- sition .
The hard glasses having electrical resistance are characterised by their negative susceptibility which may be as high as C.G.S. units .
On the other hand , the low resistance may have positive susceptibility as high as C.G.S. units .
Determination of magnetic susceptibility promises to be of some vahle in the investigation of glasses , and possibly of some of the raw materials used in glass manufacture .
There is evidence that the method would give rapidly Prof. E. Wilson .
The Measurement of the relative amounts of ferrous iron present in samples of glass , and results obtained with glasses containing known amounts of ferrous iron could be tabulated so as to obtain the actual amount of ferrous iron in any glass .
It is , of course , possible with further experiments that magnetic susceptibility may be found in glasses free from iron .
The influence of manganese and many other constituents has yet to be investigated .
21 .
Note on the Susceptibility of Tourmatine .
It well known that the magnetic susceptibility of certain crystals is a variable depending upon the direction in which it is measured .
Weiss*has examined crystals of magnetite and has determined the absolute values of the susceptibility in terms of direction and intensity of magnetising force .
He has shown that the susceptibility for moderate values of the force differs according to direction , and that the ultimate susceptibility in any direction tends to the same value when the force is sufficiently large .
The author had an opportunity of examining the magnetic properties of four specimens of tourmaline .
A good specimen of green .
tourmaline was lent by Dr. T. B. Merton .
It is in the form of a rectangular plate cm .
thick and measures cm .
in the direction of its principal crystallographic axis , and cm .
at right angles thereto .
When tested in the balance in these two directions it was found that the susceptibility at right angles to the principal axis was greater than in the direction of the principal axis in the ratio .
The dimensions of this specimen precluded an accurate statement as to the absolute value of the susceptibility .
Dr. W. T. Gordon has prepared from his collection three specimens each in the form of a cube .
Two of these were cut from a large crystal , and described as being dark blue opaque .
When tested in the balance both these specimens showed same effect , namely , that the susceptibility in the direction of the principal axis is smaller than in the other two directions at right angles to it .
The ratio of the susceptibility at right angles to the principal axis to the susceptibiliLy in the direction of this axis is and in the two cases .
The larger values of the susceptibility in the case of the smaller cube is partly attributable to its higher density , and partly to its smaller dimensions .
The third specimen was cut from a rose coloured crystal ( rubellite ) , and was too feebly magnetic to be accurately measured in the balance with the strong suspension .
A number of complete crystals were roughly tested , some of which were a light green colour at one end and rose colour at the othel .
It was evident that in all cases the green 'Jour .
de Phys 3rd Series , , p. 435 ( 1896 ) .
Magnetic Susceptibilities of Low Order .
coloured end was the more magnetic The results of the experiments are embodied in the following Table:\mdash ; Principal crystallographio axis .
22 .
General Note on the Susceptibility of Rock Specimens in Relation to the Fidd of the Earth .
A series of reports on a detailed survey of the susceptibilities of rock specimens dated November 12 , 1917 , December 17 , 1917 , January 18 , 1918 , and February 12 and 27 , 1918 , was made by the author to Sir A. Strahan , F.R.S. , Director of the Geological Survey and Museum .
They may be referred to for more detailed results obtained with the present instrument , and the work therein described has the advantage that the specimens mainly come from one particular district ( Melton ) in which magne ic observations have been made by Dr. G. W. Walker , F.B.S. A re-testing of some of the specimens therein dealt with has since been performed , and as in some cases the re-testing has been perfornled " " after demagnetisation some explanation of this process and some remarks which it calls for are desirable .
It is ] known that if a- piece of iron be subjected to a considerable magnetising force , and then tested for permeability at lower forces , the permeability then found is smaller than would have been found had the specimen been first demagnetised .
The complete previous netic history of rocks is unknown , but some of them owing to their retentivity show that they must have acquired a high degree of magnetisation at some time .
By subjecting such a specimen to an alternating magnetic orce of gradually diminishing magnitude until it ultimately vanishes , the magnetic effects due to this retentivity which have perhaps existed for 454 The Measurement of Magnetic Susceptibilities of Low Order. .
centuries can be removed in a few minutes , and it is then found that .
specimen example , dolerite may have an increased susceptibility .
In connection with magnetic survey work generally , this .
retentivity and previous magnetic history appeal to play an important part , .
because the existing magnetic effects of certain rocks are not obviously connected with the existing magnetic field of the earth .
The low grade ores example , are typical of those rocks which have a small or negligible retentivity , but certain types of rocks in which magnetite occurs have pronounced retentivity .
It would seem , therefore , that the magnetic effects exhibited by these low grade ores may be due to the existing earth 's field , or to the joint action of this field and the retained magnetism of the more basic rocks .
The subject is one of great complexity , and it does not .
appear to be satisfactory in all cases to attribute magnetic observations solely to the effects of the existing general field of the earth .
23 .
Summary .
( 1 ) The resent paper contains an account of an instrument which has been designed for the measurement of magnetic susceptibility of low order , together with some of the results to which it has led .
It depends for its action.upon the pull exerted by an electromagnet in accordance with the well-known Maxwell expression for the mechanical force exerted upon unit volume of the substance .
This mechanical force is balanced against the force of torsion in a phosphor-bronze strip .
( 2 ) A considerable portion of the paper is devoted to the determination of the instrumental constant , from data obtained directly with the itself , and by the employment of substances whose susceptibilities had been measured by other methods .
In this connection a modified method of using a ballistic galvanometer has been devised which leads to greater sensitiveness .
In addition to rock specimens , other substances have been used , and in their employment some interesting results have been obtained .
For example , it is shown that the susceptibility of 13 per cent. manganese alloy is much smaller than is usually supposed .
( 3 ) The susceptibility of powdered rock specimens has been measured and compared with that of the solid .
A very fair agreement has been obtained between the two , and the method has the advantage that powders can be rapidly made .
Care has , however , to be taken in their preparation .
( 4 ) The susceptibilities of varieties of mica have been measured , and it is shown that in certain cases in any direction parallel with the laminae the susceptibility may be over fiftyfold that obtained in a direction at right angles thereto .
Oscillations by Three-electrode Thermionic Tubes .
455 ( 5 ) A series of light aluminium alloys has been tested with the balance , and it has been found that , whereas the susceptibility of commercial aluminium is increased by alloying with copper and manganese , it is diminished by alloyin with cobalt .
( 6 ) In a note on the susceptibility of glass in relation to composition it is shown that the balance could be used to rapidly determine the relative amounts of ferrous iron in different specimens of glass .
( 7 ) Certain specimens of tourmaline have been examined .
The green and dark blue opaque varieties susceptibilities in the direction of the principal crystallographic axis varying from 16 to 20 per cenf ) .
less than in a direction at right angles thereto .
The susceptibility of rose-coloured tourmaline is very small in comparison .
( 8 ) The paper coucludes with a ote on the retentivity of rock specimens and its possible influence upon magnetic disturbances .
in magnetic survey work .
On the Variations of Wave-length of the Oscillations generated by Three-electrode Thermionic Tubes due to Changes in Filament Current , Plute Voltage , , or Coupling .
By W. H. ECCLES , D.Sc .
, and J. H. VINCENT , M.A. , D.Sc .
( Communicated by Prof. W. H. Brag , C.B.E. , F.R.S. Received November 20 , 1919 .
It is well known that a circuit containing a condenser and inductance coil can be maintained in oscillation at a frequency near its natural frequency by aid of three-electrode thermionic vacuum-tube and suitably connected batteries .
It is also known that although the frequency of the oscillations depends mainly upon the magnitudes of the inductance and the electrical capacity it is also affected by the resistance in the oscillatory circuit , by the voltages of the various batteries in use , by the temperature of the filament supplying the electrons , by other properties of the vacuum tube , and by the coupling between portions of the circuit associated with the grid and the anode .
* It is again well known that when one such vibrating circuit is caused to induce current in an independent oscillating circuit the induced current beats * See , for example , a paper by W. H. Eccles , " " Vector Diagrams of some Oscillatory Circuits used with Thermionic Tubes ' Proceedings of the Physical Society of London , ' , Part 3 , Apri116 , 1919 .

