]\gt ; II .
The Theory of Voltaic Action.\mdash ; Paper No. III .
By Professors W. E. and JOHN PERRY .
Communicated by Dr. C. W. SIEMENS , F.R. Received February , \mdash ; Read March , 1879 .
[ PLATES 4-5 .
] UP to the commencement of 1876 direct experiments on the differerlce of potentials of solids in contact with liquids had been made by VOLTA Sir HUMPHRY DAVY , MM .
BUFF , BECQUEREL , S PE'CLET , Sir WILLIAM THOMSON , and GERLAND , and of liquids in contact with liquids by KOHLRAUSCH ; SS but although extremely interesting they were all incomplete , for the following reasons:\mdash ; 1 .
The earlier experiments were not carried out with apparatus susceptible of giving accurate results .
2 .
Owing to the incompleteness of the apparatus , assumptions had to be made not justified by the experiments .
3 .
No direct had been performed to determine the difference of potential of two liquids in contact , with the exception of a by KOHLRAUSCH , using a method which appeared to us quite inadmissible as regards accuracy of result With regard to the objection No. 1 , it is needless to do more than to remark that some of the earlier experimenters thought it permissible to touch a liquid under test with the fingers , or assumed without experiment that a piece of blotting-paper , previously soaked in a liquid , acted inductively on a distant metallic surface as a layer of the liquid would .
, Gehler 's Worterbuch , Bd .
DAVT , Ann. de Chim .
, .
lxiii .
, p. 230 , 1807 .
BUFF , Ann. .
Chim . .
Pharm. , Bd .
xlii .
, , 1842 ; .
Bd .
, 1844 .
S BECQUEREL , Ann. de Chim .
et Phys. , , p. 405 , 1824 ; Pogg .
Ann. , Bd .
ii .
, , Ann. de Chim .
et Phys. [ 3 ] , .
ii .
, p. 233 ; Arch. de , p. 621 , 1841 .
PFAFF , Pogg .
Ann. , Bd .
li .
, und 197 , 1840 .
THOMSON , Proc. Lit. and Phil. Soc. of Manchester , Jan. 21 , 1862 : Papers on Elec .
and , S4UU , p. 317 .
HANKEL , Elektr .
Untersuchungen , .
; Abhandl .
der Math. Qssc dcr ) selIschaft , Bd .
, 1865 . .
GERLAND , Pogg .
Ann. , Bd .
cxxxiii .
, , 1868 .
SS KOHLRAUSCH , Pogg .
, Bd .
lxxix .
, Regarding objection No. 2 , great yagueness existed as to whether the contact difference of potentials between two nces , A and , which we shall call for brevity , was a constant depending only on the substances A and and the temperature , or whether , one or both being a liquid , it was a variable dependent upon what other substance was in contact with either .
Some writers garded it as a variable ; he had proved it constant when one or both of the substances was a solid .
But , first , the agreement of the value of the electro-motive force of each of his cells with the algebraical sum of the separate differences of potential at the various surfaces of separation , which was the test of the accuracy of his theory , is so striking , and so much greater than polarisation , \amp ; c. , usually allows one to obtain in experiments of such delicacy , one cannot help feeling doubtful regarding his conclusions ; secondly , his apparatus did not allow of his experimenting with two liquids in contact , consequently he could not legitimately draw any conclusion in this latter case .
And although KOHLRAUSCH had made some few experiments on the difference of potentials of liquids in contact , still , since he employed moist blotting-paper surfaces instead of the surfaces of the liquids themselves , we felt , for this reason if for no other , that his results did not carry the conviction the distinguished position of the experimenter might have led us to anticipate .
In fact , the inability of all experimenters to directly and with their apparatus the contact difference of potentials of a solid and a liquid , and especially of two liquids , caused the whole theory of voltaic action to remain in a vague and incomplete state .
It therefore appeared to us , in 1875 , desirable to design an apparatus and carry out a series of experiments on this subject ; and the results obtained during the following year will be found in our papers Nos. I. and II .
( Proc. Roy .
Soc. , No. 186 , .
Our method consisted in measuring directly the difference of potentials in volts at each separate contact of dissimilar substances in the ordinary galvanic cells , and then ascertaining whether the algebraical sum of all the contact differences of potential was or was not equal to the electro-motive force of the particular cell in question .
The result obtained was that , within the limits of experiment , if \amp ; c. , were separately measured ( any one or more of the substances being solid or liquid ) , then if any number .
werejoined together , and the ] motive force of the combination measured , the following equation was found true :\mdash ; Which proved that each surface of separation produced its effect independently of any other . .
JENKIN , Elec .
and Mag. , p. 44 , 1873 GERTAND , Pogg .
Ann. , Bd .
cxxxvii .
, , 1869 .
, as previously defined , being the difference of potentials between A and in contact with one another , but neither in contact with any other conductor .
CONTACT THEORY VOLTAIC ACTION .
Since a preliminary notice of our first paper was comnlunicated by Sir THOMSON to the British Association in 1876 , Professor CLIFTON has published one " " On the DifIerence of Potential by the Contact of Different Substances ; \ldquo ; but in spite of the exact workmanship of Professor CLIFTON'S apparatus and the probable accuracy of the results attainable with it , it possesses the same objection as the instrument employed by GERLAND , namely , that it was not possible to measure directly with it the rence of potential of two liquids in contact .
The results given by Professor CLIFTON axe , as far as they go , generally in accord with those contained in our paper .
II .
The experiments of which the results form the subject of the present colnmunication were made during 1877 and 1878 , and the new pparatus employed , although generally the same in principle as that previously used , possessed many important modifications suggested by experience .
The method of measurement was as follows :\mdash ; Let.3 and 4 be two insulated giltbrass plates connected with the electrodes of a delicate quadrant electrometer .
Let 1 under 3 and 2 under 4 be the surfaces whose contact difference of potential is to be measured .
3 and 4 are first connected togetl ] and then insulated , but remain connected with their respective electrometer quadrants .
alJd 2 made to change places with one another , ] being now umder 4 and 2 under 3 , then the deflection of the electrometel needle will give a measure of the difference of potentials between 1 and 2 .
Fig. 1 .
For before reversal let ( fig. 1 ) the potential of the ometer needle be , of the induction plates and of the quadrants be , of the substance 1 be , and of { he substance 2 be , so that is what we desire to measure ; * Proc. Roy .
Soc. , 1877 , vol. xxvi .
, p. 299 .
It need hardly be noticed that an tibc and not a cojlducti:e lnctlod of ) necessarily be employed since the latter old introduce always ( lil potential .
PROFESSORS W. E. AyRTON AND J. PERRY Let the of each pair of quadrants be , of 3 be , and of 4 be ; Let the coefficient of induction of the electrometer needle on each pair of quadrants be , of substance 1 on plate 3 be ) of substance 2 on plate 4 be Fig. 2 .
Now on leversal ( fig. 2 ) the needle will be deflected and the change of the capacity of each pair of electlometer ants , as well as the change of the coefficient of induction of the needle on each , will be proportional to the deflection , that is , proportional to the rence of potentials now existing between the two sets of quadrants , therefore let the potential of 3 become , and of 4 become ; Let the capacity of each pair of quadrants become , and of 3 become of 4 become ; Let the coefficient of induction of the electrometer needle on each pair of quadrants become , of substance 2 on plate 3 become , and of substance 1 on plate 4 become ; then\mdash ; ON THE CONTACT THEORY OF VOLTAIC ACTION .
This is the most general case , it only assumes the electrometer quadrants to be symmetrical with reference to the needle , but allows any amount of unsymmetry in our induction apparatus .
In practice , the motion of the electrometer needle is usually in these experiments so small that we may neglect the change in the capacity of the quadrants , and in the coefficient of induction of the needle on them produced by this motion ; consequently since the charge in the plate 3 and the attached pair of quadrants is not altered by the reversal similarly In order therefore that , which is the difference of potentials we desire to measure , shall be proportional to , the observed difference of potentials , it is necessal.y that either or and either or and either or Now the conditions on the left hand side mean perfect symmetry of our induction apparatus before and after reversal .
But as this , at any rate in Japan , is practically unattainable , it is desirable to make , and A each nought .
can be made nought and also very small by not merely connecting plates 3 and 4 together at the commencement of each experiment , but also with the outside of the electronleter or the earth , and we can also make A small by connecting either substance 1 or 2 also with the earth .
Such connexions were always made in the following experiments .
In order to rotate the lower substances 1 and 2 , it is necessary , if one or both of them be a liquid , to first increase the distance between 1 and 2 and between 3 and 4 , in order that 3 and 4 may not strike against the sides of the vessels containing the liquids .
In our former apparatus this was done by the table carrying 1 and 2 supported and capable of revolution on a central leg , which itself was movable up and down ( Proc. Roy .
Soc. , No. 186 , vol. xxvii .
, pl. 7 ) .
But when the two substances of which it was desired to measure the difference of potentials were , say , a vessel of mercury and a comparatively thin plate of metal , the difference of weight on the two sides made it almost impossible with this method to obtain sufficient parallelism of the surfaces 1 , 2 , and 3 , 4 ; so that our former experiments were necessal.ily confined to pairs of substances not having a great difference in weight .
But , as it obviously PRO SSOR q W. AYRTON AND J. PERRY desirable that the present investigation should not be thus limited , it became necessary to construct a new apparatus , of which the end elevation is shown in Plate 5 , and the side elevation in Plate 4 .
Again , in the old apparatus it was necessary to open the metal cover enclosing the whole apparatus in order to lower , rotate , and raise plates 1 and 2 .
Now we consider that , although theoretically there was no objection to opening this case , provided it was shut again to screen the apparatus from exterior electrical induction before a reading was made , still that , as it introduced a possibility of the plates 3 and 4 being touched during am experiment , this opening was objectionable .
the present arrangement , the table A , which carries the levelling screws for supporting the substances of which the contact differences of potential are to be measured ( in our figure a liquid , , and a solid plate , , about 530 square centimetres in area ) , runs with three wheels , , on a circular very rigid metallic horizontal railway , ; and to avoid lateral motion the table is kept centred by a stout iron pin , , turning in a brass socket , S. The necessary vertical motion is produced in apparatus by the ) lates anrJ 4 , which are attached to upper framework with an arrangement similar to that employed with a parallel ruler .
being raised or lowered by a vertical brass rod , capable of rotation , so as to allow of the pin , , passing through the slot , ( shown in plan below ) , or being stopped by it when the rod is required to remain raised .
The plates 3 and 4 , which are each about 300 square centimetres in area , are supported by chemically clean glass rods , , kept artificially dry by pumice-stone soaked in sulphuric acid , and resting in leaden cups , , sliding into the brass cylindrical tubes , , with bayonet joints .
These leaden cups may be lowered by handles , , so as to rest on the brass plates 3 and 4 when the instrument is not in use , thus preventing access to of the external air .
The permanent adjustment for coincidence of the plane surfaces of the plates is effected by means of three screws , , in each of the brass caps , C. Temporary adjustment of the horizontality of the brass plates is effected by means of a spirit level and the three large leyelling screws , , which support the whole apparatus , and the positions of which are fixed by hole , slot , and plane .
are strips of brass soldered to the induction plates 3 and 4 , and connected by means of clips to fine platinum wires leading to the insulated electrodes of the quadrant electrometer , and to the insulated terminals of a highly insulated short circuit key , not shown in the drawing .
In addition to the whole framework being made vel.y stout and of cherry-wood , one of the Japanese woods that warps least , it was further strengthened by various cross tiepieces , as shown in the drawings .
The whole apparatus , including the short circuit key and electrometer , was , to avoid induction outside , enclosed in a large zinc case connected with tho earth , and was not opened at all during one complete experiment , consisting of some ten short circuitings of the upper plates , reversals of the table A , and corresponding readings to the right and left of the electrometer needle .
ON CONTACT OF yOLTATC ACTTON .
We shall describe a ) operation to obtain the contact diffel.ence of potentials , between a metal and liquid , for example .
Suppose the nent a ments to have been made , and the gilt plates 3 and 4 to be quite bright .
The plate is cleaned with emery paper that has touched other metal , and all traces of the emery removed by means of a clean cloth ; it is then placed on the three levelling screws , , and fixed in position by hole , slot , and plane .
The porcelain dish containing the liquid is laid in a metal onejust fitting it , and on the base of which is a hole , slot , and plane ; this is now laid on the other levelling screws , The rod , , is then lowered umtil the disk , , rests on a plate let into the top of the wooden frcunework at the top of the -that is , until the induction plates and 4 are in their lowest position .
The levelling screws , now raised until a small metal ball , of a diameter of 8 millims. , is in contact at three fixed points with the plate 4 and the plate , or until when in contact with the plate 3 it and its reflection in the liquid appear to meet .
To avoid any harm arising from possible contact of the liquid with this gauge ball , it was made of a material not acted on by the particular liquid under experiment .
Before proceeding , each pair of quadrants is in succession put to earth , the other pair remaining insulated in order to test for any possible from the needles to the quadrants .
Each pair of quadrants is now charged with a battery , the other pair being connected with the earth , in order to test for any along the glass rods , the small glass rods supporting the quadrants in the electrometer , or along the paraffined ebonite pillars of the short circniting key .
It having thus been ascertained that there is no leakage , the strip of metal which has been cut from the same sheet of metal as itself , and attached to it by a binding screw soldered to , is made quite bright with enlel.y paper and a cloth , and its end dipped into the liquid , as shown in Plate 4 .
The zinc case is then closed up , plates 3 and 4 connected together , and with the earth , by means of a key ( the handle of which was a long thin ebonite rod projecting through the zinc case ) , and the electrometer reading taken .
3 and 4 are then insulated from one another , and from the earth , and raised by means of the rod , , projecting above through the zinc case ; the ] A turned from below by means of a handle ssing through the base of the instrument ; 3 and 4 then lowered into exactly their former position , ensured by the parallel motion of the supporting beam and by the limiting stop , .
The reading of electrometer is now taken .
Again short circuit , insulate , raise , reverse , and lower , and take a new electrometer reading , \amp ; c. Some ten readings having been thus obtained , a fresh set of expel.iments is always made with the same two substances in the following way in order to compensate for the error introduced by defects in parallelism of the apparatus affecting the obtained from two rigid surfaces ( as that of copper and , differently result found with one or with two liquid surfaces under test .
Instead of before , with the liquid under 3 and the plate under 4 , we start with the plate under 3 and the liquid under 4 , and readjust , by means of the levelling screws , the heights of the surfaces , until their distance from the plates 3 and 4 is , as before , 8 millims. We then short , insulate , l.aise , reverse , and lower , and take exactly as many readings as before ; and the mean of the two sets of readings , obtained with the wo modes of levelling , is regarded as the result of the particular experiment .
III .
The results we have already obtained in this present investigation group themselves under three 1st .
The contact difference of potentials of metals and liquids at the same temperature .
2nd .
The contact difference of potentials of metals and liquids when one of the substances is at a different temperature from the other in contact with it ; for example , mercury at C. in contact with mercury at C. 3rd .
The corjtact difference of potentials of carbon and of platinum with water , and with weak and strong sulphuric acid .
But those contained under head No. 1 are alone contained in the present investigation .
The remainder of our completed experiments we desire to have the honour of submitting on a subsequent occasion to the Royal Society .
A fourth very important division of the subject , namely , the measurement of the contact difference of potential of substances in other gases than air , it has been our intention for some time to take in hand ; but , although we commenced the working drawings of the apparatus necessary for that investigation at the beginning of 1877 , we have not yet had an opportunity of constructing it .
The numbers which follow in the accompanying tables are those to which alone we attach importance , but it need hardly be said that , in such delicate experiments as the present , much time has to be spent in obtaining measurements which are afterwards found out to be wrong for a variety of reasons .
All such we have not included in the following tables ; and this therefore explains why , in some cases , one measurement only is apparently the result of a whole day 's work .
In our previous papers Nos. I. and II .
we used as our standard electro-motive force that of a LATIMER CLARK'S standard mercurous-sulphate cell , since that was the most constant galvanic cell known to us ; and we found experimenta ] lie , as Sir WILLIAM THOMSON previously ascertained , that zinc had to copper a contact difference of potentials volt .
Now , although .
LATIMER CLARK'S cell is a very constant element , the difference of potentials of zinc and copper is even more constant ; we have , therefore , in the present investigation compared all our measurements with this standard , which we have taken to be equal to volt .
ON THE CONTACT THEORY OF VOLTAIC ACTION .
We now endeavoured to measure the contact differences of potential of each pair of constituents in a LATIMER CLARK'S mercurous sulphate cell .
First we tried mercury and mercurous sulphate paste , the mercury being put into one of our porcelain dishes and the mercurous sulphate paste into the other , the connexion between the two being made by a small glass siphon filled with the mercurous sulphate paste nd the earth connexion by a platinum wire dipping into the paste in the porcelain dish .
Many days were spent unsuccessfully at this experiment , and at similar ones which follow , on account of the thin layer of water which floated on the top of the mercurous sulphate , even after great care had been taken to remove it with a pipette .
Often , therefore , instead of measuring the contact difference of potentials between a substance A and the mercurous sulphate paste , we obtained that of A with the paste plus that of the paste with water .
At length , however , we obtained the following , each nnmber given being , as usual , the mean of many experiments .
Very great care was taken to have the materials chemically pure , the mercury frequently redistilled ; also to prevent any discrepancies in these results being caused by any possible tion of the porcelain vessels themselves , they were thoroughly de-electrified after cleaning and before using .
It may be noticed that this difficulty ienced by the thin layer of which floated on the mercurous sulphate paste is a very good example of the inaccuritcies that must have been introduced by former experimenters using a moist blotting paper surface instead of the surface of a liquid See page 15 , " " Reasons for Incompleteness of Early ters In order to ascertain whether the high or the low yalue for the constant difference of potentials of zinc sulphate with merourous sulphate was correct , an incomplete cell of zinc sulphate solution , mercurous sulphate paste , and mercury was formed , and the electro-motive force measured , arrangement employed consisting of the zinc sulphate solution being put in one of the porcelain dishes , and in the other the mercury , the two being connected with a small beaker containing mercurous sulphate paste , and which did not act inductively on the induction plates .
The connexion between the zlnc sulphate and the mercurous sulphate was made by a clean cotton wick previously soaked in a saturated solution of the former , and the connexion between the and the paste in the beaker by a small glass siphon tube filled with the latter .
Earth .
connexion in the first experiment recorded was made by a platinum wlre nto the mercury , afterwards by a platinum wire dipping into the zinc sulphate lnstead of into the mercury .
SATURATED zinc sulphate and mercury connected by mercurous sulphate paste .
ON THE CONTACT THEORY OF VOLTAIC ACTION .
Now the mean of the low set of values obtained previously with zinc sulph te and mercurous sulphate is , and of the high values ; the mean of the values previously obtained with mercury and mercurous sulphate is , and a single value .
In order to obtain the electro-motive force of the incomplete cell we must subtract the two latter from the two former , giving us or which are practically equal to the extreme values obtained directly for the electromotive force of the incomplete cell .
It therefore seems to be tolerably certain that the variable results obtained when experimenting with mercurous sulphate paste are due to the uncertainty of the nature of the bounding surface .
The mercurous sulphate paste and the saturated zinc sulphate solution were now boiled together , as is done in preparing Mr. LATIMER CLARK'S constant element .
Using this mixture in one porcelain vessel and mercury in the other , and them with a small glass siphon filled with the mixture , we obtained Volts .
Mean of .
9/ 5/ 78 Mercury has to C. 10 11/ 5/ 78 , , , , , , 10 Wetting the the mixture with a little saturated zinc sulphate solution up the answer to , the latter being the mean of ten observations all giving nearly equal results .
The following contact difference of potentials no measured .
yolts .
11/ 5/ 78 Amalgamated zinc has to Mixture , A little saturated zinc sulpha solution being poured over the mixture and taken out again carefully with a pipette altered the result to The following incomplete cell consisting of mercury and amalgamated zinc joined by the mixture of saturated zinc dphate and mercurous sulphate paste previously referred to was now made , and the mean electro-motive force was found to be about yolt , the mercury being positive to the ama]gamated zinc ; this is roughly equal to the first value obtained with mercury and the mixture added to the last value obtained with the mixture and amalgamated zinc .
It was found that yolts .
26/ 4/ 78 Amalgamated zinc had to Copper C. From which it follows that Amalgamated zinc had to Platinum , , Therefore the electro-motive force of a LATIMER CLARK'S cell should be Volts .
( Mercury ) Mercury , Amalgamated zinc ) Platinum ) 1.481 The real value of the electro-motive force of the cell being about volts .
The connexion between the liquids was made by a glass rod dipped previously into tilled water , and the legs of which rested one in each of the porcelaiJl dishes .
During each of the experiments the readings gradually rose , but if the glass rods were taken out , redipped in distilled cvat , , and replaced , they returned to their original value .
In case this might be due to the drying up of the film of water on the surface of the connecting glass rod , the connexion between the liquids after the next experimade in the following way .
In addition to the wetted glass rod , the legs ON THE CONTACT THEORY OF VOLTAIC ACTION .
of which dipped into the two porcelain dishes , there were twisted round it a few clean cotton threads , which had been previously soaked in distilled water or in the non-acid liquid employed , alum , copper sulphate , \amp ; c. The threads were of such a length that one of their ends dipped into the dish containing the non-acid liquid , while the other descended nearly , but not quite , to the acid .
V. As a test of the accuracy of the measurements we may see how nearly for any three metals , the sum equals nought , as it ought to , since it is well known that in any compound metallic circuit at umiform temperature there is no electro-motive force .
In order that the test may be the more rigid we will employ only the measured contact differences of potential , and not the calculated ones or those marked with an asterisk in the table of contact differences of potential of solids with solids .
Now if it be remembered that the various experiments with the different metals , although made at an average temperature of C. , were , as mentioned in the various tables , made at somewhat different temperatures the one from the , it wiIl be seen that the previous six sums cannot be absolutely llought .
We may therefore regard them as sufficiently small to prove a considerable amount of accuracy in the experiments .
The numbers given in the two large tables for the contact differences of potential of solids with solids , solids with liquids , and liquids with liquids might be employed to build up the electro-motive forces of well known one and two fluid galvanic cells .
As however the summation law referred to at the commencement of this paper may be considered as having been established in our two previous papers on the contact theory of voltaic action , it is not necessary in the present instance to take further examples of it .
In all our experiments two air contacts enter our measurements , and the summation law gives no indication of the value of the contact difference of potentials at these air contacts .
For example , suppose it be desired to measure the differeI ] of potentials between two substances A and in contact , either or both of which msy be solid or liquid , then in reality what we measure is .
If now we measure by the inductive method the difference of potentials between and in contact , then what we really obtain is ( Air ) , and the sum of these two we have proved in all cases to be equal to what is obtained if A and be joined by and the difference of potentials between A and measured .
What really we have proved is therefore that .
And this would be the case whatever the values of the difference of potentials at the air contacts , provided they remained constant , so that .
One way , and the way we hope shortly to employ , to enable us to determine the electro-motive force of contact , in volts , of a substance and a gas is by repeating all these exact contact experiments in different gases , for if we measure by our inductive method contact difference of potentials of A and in a gas then we shall obtain ON THE CONTACT THEORY OF VOLTAIC ACTION .
and the difference between this and the result of the first measurement will be Consequently , if such experiments be made and the exact electro-motive forces of contact measured in a number of gases , we shall ascertain whether the contact difference of potentials of a substance and a gas differs much with different gases .
Qualitative experiments in this direction have already been made with very interesting results by Mr. BROWN , of Belfast , but his experiments differ from ours in not being quantitative .
If the gas measurements such as we have indicated be extended to a good CROOKES ' vacuum we may possibly approximate to the real value of , a contact difference of potentials of A with , the value that we should obtain by a measurement of the PELTIER effect .
It is therefore our purpcse to extend these quantitative ments not only to other gases than air but to a CROOKES ' vacuum , as well as to measure the PELTIER effects for a nllmber of acts of dissimilar substances .
We have to thank two of the students of telegraphic engineeling at the Imperial College of Engineering , Japan , Messrs. T. IWATA and N. , for considerable assistance rendered us in the carrying out of this somewhat 1borious investigation .

