]\gt ; .
Measurements of Specific Inductive Capacity of Dielectrv , in the Physical Laboratory of the University of lasgow .
By JOHN C. GIBSOI , and BARCLAY , .
Communicated by Sir WILLIAM } , F.R.S. Received November 23 , 18 \mdash ; Read February 2 , 1871 .
THE object of this paper is to describe instruments and processes employed in a series of experiments on the specific inductive capacity of paraffine , and the effect upon it of variations of temperature .
The chief instruments employed in this investigation were the Quadrant Electrometer , and two others which will now be described .
I. The Platymeter , so named by suggestion of the late Dr. from its use in the comparison of electrostatic capacities , was , in a rudimentary form , shown to the Mathematical and Physical Section of the British Association at its Glasgow Meeting in 1855 , by Sir .
It consists of two condensers of equal capacity , each of which has one of its opposed surfaces in metallic connexion with the corresponding surface in the other .
In the instrument employed ( see Plate XLIII .
fig. 1 ) the conductor constituting these two connected surfaces is a cylinder centimetres in length and centimetres in diameter , supported at its ends in a horizontal position by two vulcanite stems .
Round it are placed two equal and similar pieces of brass tubing , centimetres in length and centimetres in diameter , each supported by a vulcanite stem , so as to be concentric with the cylinder and at equal distances from its ends .
The four supports are fixed in a massive plate of iron ; and the whole is enclosed in a stout metal box , an electrode from the inner cylinder being carried through an insulating plug ( o ) of paraffine in the end of the box , and another from each of the tubes through similar plugs the top of it .
By this means shreds and dust are excluded , and paraffine is found to be so good an insulator that no inconvenience is experienced by employing it in this way .
II .
The Sliding Condenser is a condenser the capacity of which may be valied by altering the effective area of the opposed surfaces .
The following form of this instrument was employed ( see fig. 2 ) .
Two pieces of brass tubing , centimetres in diameter and and centimetres in length respectively , are supported end to end upon vulcanite stems securely fixed in a massive iron plate Within these tubes , and coaxially with them , moves a brass cylinder centimetres iu length centimetre diameter , supported at one end upon four brass feet which slide on the inner surface of the longer tube , and loaded at that end so as to rest stably upon them .
Attached to this cylinder is a vertical arm ( g ) which projects through a slot extending along the upper part of the surrounding tube .
This arm carries an index ( i ) by which the position of the sliding cylinder or core is indicated on an attached scale ( It ) .
This scale is centimetres long , and is divided into 360 divisions , each of an inch in length .
The tube which supports these parts of the instrument , though placed on vulcanite stems for convenience in testing , is in working kept in connexion with the earth .
The other , which may therefore be called the insulated tube , is closely surrounded by a sheet of metal , which is securely fastened to the iron base of the instrument .
This sheet , being connected with the earth , guards the insulated tube against electric disturbance , and adds largely to the capacity of the condenser .
As an additional security against disturbance , another sheet of metal is fastened round the former at some distance from it .
In using these instruments the following connexions are made ( see Plate XLIV .
fig. 3 ) .
The inner cylinder or core of the platymeter is connected with two of the quadrants of the electrometer , the other two being connected with the case of the electrometer and with the earth .
The condenser to be measured is connected with one of the tubes or sides of the platymeter , and the insulated tube of the sliding condenser with the other .
All other parts of the sliding condenser and the box and base-plate of the platymeter are kept in metallic communication with one another , and with the disinsulated pair of quadrants of the electrometer .
Now suppose the two sides ( and ) of the platymeter to be exactly equal , and the condensers to be compared , say A and , also equal .
Let the insulated quadrants be temporarily connected with the disinsulated pair and let the condenser together with , the side of the platymeter wit'n which it is connected , receive a positive charge .
Then , the core of the platymeter , becomes by induction negatively electrified .
The earth-connexion of the quadrants is now broken , and the charge thus insulated upon remains masked by the action of so as to cause no deflection of the needle .
connexion is now established between and , the two sides of the platymeter , so that on the supposition of equality already made , the charge is now distributed over a system of double capacity , and the potential of the whole becomes half of that of A and before distribution .
Both sides of the platymeter are now acting upon each with half the original intensity of , so that the resulting effect upon the core being unaltered , no deilection of the needle takes place .
If the condenser A be of greater capacity than the condenser , then , on connecting and , the capacity of the whole charged system is less than double that of A and and its potential , therefore , greater than half the original potential of A and .
Each side of the platymeter , therefore , is now acting with more than half the original intensity of .
The result is an increased action on the core of the platymeter , and the consequent liberation of positive electricity to act on the needle , which , in the ordinary arrangement of } electrometer , is deflected to the right .
If , on the other hand , the capacity of the condenser A be less than that of the conSPECIFIC INDUCTIVE CAPACITY OF DIELECTRICS .
denser B , the potential of the system after the distribution of the charge is less than half the original potential of A and .
The combined action of the two sides and is now less than the original action of the one side .
Negative electricity is thus set free , and this , acting on the needle , deflects it to the left .
If be the charged condenser , or if A be charged negatively , the effect in the last two cases will be reversed .
If both the condensers be charged , one positively and the other negatively , the effect will be greater in degree , but precisely similar in kind .
This is the method usually adopted , as it gives more marked indications without increasing the risk of failure of insulation .
In measuring any condenser by comparing it in this way with the sliding condenser , the latter is , in accordance with the indications thus given by the electrometer , so adjusted that its capacity becomes equal to that of the condenser to be measured .
In order to find from this the true value of the measured condenser , it is necessary to know the value of the sliding condenser in scale-divisions when its index is at zero on the scale .
For this purpose a condenser was employed of such a form that its capacity could be very accurately determined in absolute measure*(see fig. 4 ) .
This consists of two metallic spheres of different diameters placed one within the other .
The outer consists of two brass hemispheres ( a having their inner surfaces accurately turned , and having projecting flanges which are firmly fastened together by three brass screws ( one of which is shown at ) .
In the top of the upper hemisphere is a small hole through which projects an electrode from the inner sphere , which rests upon three vulcanite pins ( one of which is shown at ) , so as to present an accurately spherical face t ] concentric with the inner surface of the outer sphere .
In order to determine in absolute measure the electrostatic capacity of this condenser , its dimensions were calefull ascertained in the following manner .
The quantity of water contained by the outer sphere was measured , and foumd to be ains at temperature Centigrade , or cubic centimetres , which , with the necessary correction for the vulcanite pins , 287 cubic centimetre , gave as the radius centimetres .
The radius of the inner sphere was obtained from the content of the outer sphere and that of the space between the two spherical surfaces .
Of the content of this space five determinations were taken .
These are shown in the following Table .
values in the second column include the correction for the volume of the vulcanite pins .
*The absolute unit of electrostatic capacity referred to in this paper is the capacity insulated )herical conductor of one centimetre radius placed at an infinite distance from all external objects .
The mean of these values , and the previously measured volume of the contaimng spherical surface , give centimetres as the radius of the inner sphere .
The capacity of a spherical condenser is calculated by the formula Capacity ' where is the radius of the outer , and the radius of the inner sphere .
In this case centimetres and centimetres .
Hence the capacity is equal to centimetres .
The specific inductive capacity of the vulcanite of the pins which support the inner sphere being greater than that of air , causes an increase of the whole capacity of about19 centimetre .
The hole in the top causes a diminution centimetre , and the electrode passing through it an increase of 15 centimetre .
The actual value of the condenser is therefore centimetres .
To reduce this to scale-divisions of the sliding condenser , the value of one division in absolute measure is calculated by the formula for a cylindrical condenser , Capacity where is the radius of the outer surface , the radius of the inner surface , and the length of the condenser whose capacity is to be calculated .
In this case centimetres , centimetre , and of an inch or 063499 centimetre .
Here was determined from a measurement of the volume of water contained by the tube , the length of which was accurately measured .
To determine , the circumference of the core was measured by winding fine wire round it , and measuring the length of a certain number of turns , the necessary corrections being made for the thickness of the wire and its spiral arrangement .
The value for electrostatic capacity of one scale-division is therefore 0413 centimetre ; and hence the spherical condenser is equal in capacity to scale-divisions .
Direct electrical measurements taken subsequently on a sliding condenser of greater range gave 1607 scale-divisions as the value of the spherical condenser .
This is probably greater than the actual value , while that derived from calculation is too small by quantity due to the action of parts whose capacity could not be numerically determined .
The mean of these values , 1572 scale-divisions , may therefore be taken the value of this condenser .
] When the capacity of the spherical condenser was measured electrically while in connexion with the side of the platymeter , the reading obtained on the scale of the sliding condenser was 211 .
When the connexions were reversed , that is , when the spherical condenser was connected with the side of the platymeter , the reading obtained was 183 .
The difference of the two readings thus obtained shows some inequality of the sides of the platymeter .
Now to find the true reading in any case from two such readings , suppose the side to be equal to .
If no deflection of the needle takes place on the distribution of sPECIFIC INDUCTIVE CAPACITY OF DIELECTRICS .
the charge , this can only arise from the action of the two sides of the platymeter being equal to that of the one charged side , say , before the connexion was made .
But the surface of action is now equal to , and the potential of the system , is now , where is the original potential of A and .
But the quantity of electricity taken from A and is such as to raise and to the same potential as that to which A and , with which they are now connected , are reduced .
But in condensers at equal potentials the capacities are to the amounts of the charges .
Suppose the original charge of A and to be unity .
Then their charge after distri .
bution is , and that of and is .
Then ; but and therefore That is , generally , when there is no deflection , the condensers compared are to one another as the sides of the platymeter with which they are connected .
But the reading obtained with the platymeter connexions arranged normally was greater than that obtained when they wele reversed .
The side is therefore greater than the side Now let be the greater and the less reading with the value of the sliding condenser when its index is at zero added to them , and let be the true value to be deduced from them .
Then we have the ratios and : therefore , and That is , the true value of a condenser measured by an imperfect platymeter is the geometric mean of the two values obtained by the different arrangements of the connexions .
In practice , when the error of the platymeter is so small as it is in the present case , the arithmetic mean may be taken instead of the geometric .
For let be the arithmetic mean of and , and the difference between it and either of these values .
Then .
In the present case , lvhere is about does not differ sensibly from .
If be the greater and the less then , the value ( in scale-divisions ) of the sliding condenser with its index at zero , is .
When has been determined , the true value of any condenser may be got by adding to it the mean of the two readings obtained ; and when the elror of the platymeter used has been once ascertained , this mean may be obtained from a single reading without reversing the connexions .
The following Table shows the values obtained for and in the measurement of the spherical condenser:\mdash ; Taking 198 as the mean reading and 1572 as the value of , the value of the sliding condenser with its index at zero was 1374 scale-divisions .
A condenser was now prepared having paraffine stead of air as the dielectric .
This consisted of a flat circular brass box with a tube projecting from the centre of the lid , through which an electrode was carried from a brass disk imbedded in paraffine , with which the box and tube were filled , midway between the top and bottom of the box .
The box was placed in water reaching nearly up to the top of the tube .
For the value of this condenser at different temperatures the following results were obtained:\mdash ; Crevices which had formed in the paraffine .
cooling , gradually admitted water so as ultimately to destroy the insulation of the disk .
The increase of capacity thus caused is probably sufficient to account for the variations shown in the Table , which , therefore , are not to be attributed to any alteration of inductive capacity due to change of temperature .
Taking the first of the values , 1542 scale-divisions , or centimetres , as probably the most accurate , the following determination ofspecific inductive capacity ofparaffine is obtained .
The value of the box condenser with air instead of paraffine as the dielectric , obtained from measurement of it on the sliding condenser described in the Appendix , is centimetres .
These two values give as the specific inductive capacity of paraffine when that of air is taken as unity , a correction having been made on account of three vulcanite pins used to support the insulated plate .
The following form of condenser ( see fig. 5 ) was next employed as being most likely to show effects of temperature undisturbed by other causes .
Into a cylindrical brass vessel centimetres in depth and centimetres in internal diameter , melted SPECIFIC INDUCTIVE CAPACITY OF DIELECTRICS .
paraffine was poured to a depth of about one centimetre .
Upon this , when solid , piece of brass tubing , centimetres long and centimetres in internal , and in external diameter , was rested so as to be concentric with the outer tube .
Inside of this , and concentric with it , was placed another piece ofbrass tubing centimetres long and centimetres in external diameter .
The space between this and the outer tube was then filled with paraffine , and from the middle tube thus imbedded a fine wire electrode was carried up .
The condenser was put into a vessel containing water in which was placed a thermometer ; and another thermometer was supported in the centre of the inner tube by a paraffine Plug ( ff ) about a centimetre thick , which rested upon the top of this tube .
This plug prevented the communication of the temperature of the air .
As an additional security , another paraffine of the same thickness was inserted in the inner tube a little above the bulb of the thermometer .
The outer and inner tubes were connected with the earth , and the electrode from the other tube , which was insulated , was connected with the one side of the platymeter .
An addition to the capacity of the sliding condenser had been made , increasing its zero-value , so that the readings for the spherical condenser were : This gives for the zero-value of the sliding condenser scale-divisions .
With this value the results shown in the following Table were obtained for the value of the cylindrical paraffine condenser at different temperatures .
Without the addition to the value of the sliding condenser the ments of the sphencal condenser were taken:\mdash ; IIDCCCLXXI .
4 These readings give the zero value .
With this value of the ffillowing measurements of the cylindrical paraffine condenser were obtained:\mdash ; In order to vary the conditions of the experiment so as to allow the paraffine to expand more freely , a condenser of the following form was employed .
Its metallic surfaces consisted of circular pieces of tinfoil .
These were arranged horizontally with plates of paraffine half a centimetre thick between them .
The following Table shows the values obtained for this condenser at various temperatures , the zero-value of the sliding condenser being still the same : These values arranged according to temperature are as follows:\mdash ; Some experiments were now made upon the expansion of paraffine with temperature , with a view to determine the alterations produced by it in the capacity of paraffine condensers .
This was done by weighing a quantity of paraffine in thin plates , with a platinum sinker attached , in distilled water at different temperatures .
The weight of the paraffine in air was grammes , that of the platinum was grammes , or SPECIFIC INDUCTIVE DIELECTRICS .
ltogether grammes .
The following Table shows the results of the weighings in water .
The weights are given in grammes , and the volumes in cubic centimetres .
These numbers give for the expansion of paraffine the results shown in the following Table * ansion pinear e From these determinations of the expansion ofparaffine the effect produced byvaliations of temperature upon the capacity of the tinfoil condenser may be easily calculated .
The mean value of the condenser scale-divisions may be taken as its value at the mean temperature .
Its values at the different temperatures estimated from this , together with those previously given as obtained from actual measurements , are shown in the following Table:\mdash ; The values given in the last column of this Table were estimated on the supposition that the expansion of the paraffine did not produce any sensible stretchin of the tinfoil ; and this , from the manner in which the condenser was formed , was probably the case .
From this it appears that the regular alteration in the values obtained for this condenser at different temperatules follows so nearly that resulting from the variations of the * After these experiments were made it was found that the French ' Burcau des Longitudes in the 'Annuaire ' for 1870 published by them , give , on the authority of M. FIZEAU , 00027854 per degrcc ritdc ils the coefficient at Centigrade oflinear expansion of ' ' paraffine dc \ldquo ; ) ils the tion of coefficient\ldquo ; per degree .
This gives the expansion at tcml ) craturc 9 given in the at the same temperature being 000240 .
bIESSRS .
GIBSON AND BARCLAY ON MEASUREMENTS OF distance between the plates due to the expansion of the paraffine that no change of specific inductive capacity can be inferred from it .
The following further measurements of the Cylindrical Paraffine Condenser were now made , the zero-value of the sliding condenser being still equal to 1384 scale-divisions .
All the values obtained for this condenser at different temperatures are shown in the following Table arranged in the order of temperature:\mdash ;

