Sixth Letter on Voltaic Combinations .
Addressed to MICHAEL FARADAY , Esq. , D.C.L. F.R.S. , Fullerian Prof. Chem. Royal Institution , &amp; 2c .
Ssc .
&amp;c. By J. FREDERIC DANIELL , Esq. , For .
Sec. R.S. , Prof. Chem. in King 's College , London .
Received April 22 , -Read April 28 , 1842 , MY DEAR FARADAY , I MUST beg permission to address you once more upon the subject of Voltaic Combinations .
To this I am prompted by several considerations .
In the first place , the beautiful law of OHM , and the simple expression which he has given of the electromotive force and resistances of a voltaic circuit , enable me to review with advantage , and to correct , many of the conclusions which I had derived from former experiments ; and have suggested additional experiments , the results of which will tend , I trust , to remove some obscurities and ambiguities which were left in my former communications .
2nd .
By following out these principles I shall be enabled to offer some practical remarks upon the different forms of voltaic batteries which have been brought forward to assist the speculations of the active inquirers , who , in the present day , are so eagerly engaged in applying the voltaic forces to the service of the arts .
3rd .
I wish most particularly to explain more fully the principles of the cylindrical arrangements of the battery which I have introduced , and which appear to me to have been greatly misunderstood .
I am desirous , however , that you should understand that I do not present the following observations for the purpose of testing the law in question , or of determining constants connected with the formula , for that could only be satisfactorily effected by experiments of a much more delicate and accurate nature than those to which I shall have to refer ; but with a view to show how generally the law applies , even to the practical results of operations carried on upon , what might be called , a manufacturing scale , in which disturbing influences are numerous , and in a great measure uncontrolable .
Professor OHM has adopted ( I believe that you will concur with me in thinking , unfortunately ) the contact theory of the electromotive force ; and although his formula is easily adapted to either of the two rival views , it is perhaps necessary , in selecting the chemical theory , that I should define the exact meaning which I attach to his symbols , and explain the expansions which I think it necessary to introduce .
The formula , you will remember , is E R+r A , where E represents the electromotive force ( so called ) in the cell : R the resistances in the cell : r the amount of exterior resistances : A the effective force , measured by the work performed .
Now according to the chemical view , E must be the balance of several active forces in the cell .
1st .
The superior affinity of the generating plate for the anion , of the electrolyte , which we will designate by B. 2nd .
The inferior affinity of the conducting plate for the same anion , which we will call b. 3rd .
The affinity of the cation , disengaged from the electrolyte and accumulated upon the conducting plate for the anion , this we will call e ' : these two last tend to produce polarization , as it is not very appropriately called , and a current in the opposite direction to B ; therefore E=B ( b + e ' ) , or EB -b e ' .
R , the resistance in the cell , varies directly as the thickness of the electrolyte ( or the distance between the generating and conducting plates ) , D , and inversely as the area of the section of the electrolyte , S ; therefore R = .
r represents all the exterior resistances , whether metallic or electrolytic .
In the metallic parts of the circuit , it will be inversely as the area of the section ( i. e. the square of the diameter ) of the wire s , and directly as the length 1 , or distance through which the current passes .
In electrolytic work this metallic resistance may generally be disregarded ; the lengths of the connecting wires being insignificant with regard to the resistance of the electrolyte .
This latter will be inversely as the area of the section of the electrolyte , s ' , and directly as the distance d between the electrodes : therefore Id r-+7 ' r= i But we must now inquire particularly what it is we mean by the section of the electrolyte .
The limits of the section of the metallic conductor are strict and easily * The resistance of each liquid is specific , and depends upon the nature of the liquid , the degree of saturation ( if a solution ) , and the temperature ; these circumstances remaining constant , it also is constant , and is so considered in this review of the formula .
138 determined ; but , taking into consideration the diffusive nature of the electrolytic force , and the wide spread of that polarization of the particles of an electrolyte which we have traced upon a former occasion to the back surface of a conducting plate opposed to a mere point of generating metal* , it is more difficult to define the limits of its action so as to satisfy the conditions of theformula .
In a cell composed of a plate of generating metal with a conducting plate of equal dimensions , the interposed electrolyte only wetting the opposite faces of the two metals , the area of the section of the electrolyte will clearly be equal to the area of the acting surface of the conducting plate .
In case the two metals should be imimersed in a trough , in such a manner as to allow of the electrolyte being in contact with both sides of the plates , it is also probable that the action of the back surfaces might be disregarded without danger of material error in our calculations , although we know in fact that they would not be wholly passive .
Up to this point , therefore , there is no difficulty in the application of OHM 's formula .
But how are we to determine the area of the section of the electrolyte , when the surfaces of the generating and conducting plates are not equal ?
as , for instance , in the case of a rod of zinc placed within a cylinder of copper .
Is it referrible solely to the surface of the conducting plate ?
Or is it limited by the mean of the surfaces of the two plates ?
The experimental investigation of this point , although the final result is extremely simple , has cost me much labour .
The apparently unlimited spread or radiation of the force from a point of generating metal over an indefinitely large surface of conducting metal , strongly suggested the first conclusion .
This was moreover confirmed by the following consideration , viz. if we take the mean section of the electrolyte as determined by the mean of the surfaces of the two metallic plates between which it is included , it is clear that the result ought to be the same , whether the generating or conducting metal be the larger of the two .
A rod of copper placed within a cylinder of zinc , ought to circulate the same amount of force as a rod of zinc placed within a cylinder of copper ; the dimensions in both cases being respectively the same .
Now the results of a vast number of experiments , some of which I have already submitted to yout , seemed to prove that , so far from this being the case , the amount of force is reduced one-half when the lines of force are made to converge from a large generating surface towards a small conducting one ; instead of diverging in the contrary arrangement , from a small generator to a large conductor .
These experiments I have again repeated , and when I made use of sulphate of copper in dilute sulphuric acid as the electrolyte , with the same general results .
The impossibility of reconciling this with the law , and the necessity of determining a point of such fundamental importance , together with a certain degree of unsteadiness of action in the experiments , induced me , at length , to change the electrolyte for one which would be less liable to alter its condition during the progress of the Philosophical Transactions , 1838 , Part I. p. 54 .
t Ibid. , pp. 47 , 49 , 53 .
T2 139 observations .
The arrangement which I adopted was that of hollow cylinders of amalgamated zinc with platinum wire , and wire of amalgamated zinc with platinum cylinders , all of equal heights ; and the electrolyte in contact with the zinc consisted of standard dilute sulphuric acid , separated by a porous tube from strong nitric acid in contact with the platinum .
A BREGUET'S thermometer , adapted to the purposes of a galvanometer* , was selected as a measure of the effects .
The following Table exhibits the results : TABLE I. Diameter of Zinc .
Diameter of Platinum .
Degrees of Galvanometer .
inches .
24 wire 274 Mean of three observations .
wire 2 .
255 Mean of three observations .
wire 1I 279 1wire 273 270 Mean .
The needle always returned after each experiment to the point from which it started .
There can be no difficulty in taking these results as sensibly equal ; and it is therefore evident that a wire of platinum placed within a cylinder of zinc , established a current of exactly the same force , as a wire of zinc placed within a cylinder of platinum of equal diameter .
Hence we may conclude that the area of the efficient section of the electrolyte is the mean of the opposed faces of the metal plates .
But how shall we account for the contrary results with sulphate of copper as the electrolyte , which were carefully made and greatly varied , but which always gave a consistent result of about one-half the force when the con(lucting was very small in proportion to the generating surface ?
By substituting acid sulphate of copper for nitric acid in the arrangement just described , and carefully attending to the progress of the experiment , the explanation became obvious .
When the platinum wire was placed in the porous tube in the centre of the zinc cylinder , upon first connection with the calorific galvanometer , the needle advanced to 83 ?
; but almost immediately began to return till it reached 28 ?
, at which point it remained stationary .
When the wire was withdrawn it was found covered with copper in a spongy or pulverulent form .
It was wiped and replaced , and the needle advanced to 85 ?
, but immediately began to retrograde and fell again to 28 ? .
This process was repeated many times , and always with the same result .
If , while the galvanometer was at its highest point , the wire was moved about and the liquid kept in a state of agitation , the needle remained steady for a longer time ; but when the wire was left at rest , the needle always receded .
When , on the contrary , a zinc wire was placed in the porous tube in the centre of the platinum cylinder , charged with the sulphate of copper , the needle advanced to 960 upon an average of three experiments , and then remained quite steady .
In this case , the precipitated copper was equally diffused over the surface of the platinum , and constituted a compact layer firmly attached to the plate .
There can be no doubt that it is this difference in the state of the precipitated metal which gave rise to the difference in the results of the two arrangements .
In the last case , the electrolysis of the liquid was carried on without the disengagement of any element tending materially to produce an opposing current ; while in the first , the spongy state of the copper retained the liquid within its pores ; which , after the precipitation of all the sulphate of copper which it contained , generated hydrogen , which was equally entangled in it , and produced a strong opposition to the current .
The amount of this opposition is definite , and of nearly half the force of the principal current ; and hence I was led to the erroneous conclusion regarding the relative sizes of the generating and conducting plates .
Taking the measure of the force at the first moment of making the contacts , the results sufficiently confirm the conclusion drawn from the experiments with the nitric acid .
I once more tested the hypothesis , that it is the mean section of the electrolyte which regulates the current , and that it is indifferent whether the conducting or the generating metal be the larger of the two plates , by measuring the chemical results produced .
For this purpose I weighed the amalgamated zinc cylinders and rods before and after the experiment , and ascertained the consumption of metal for intervals of half an hour , during which the circuits were closed .
The conducting metal was copper , and the rods half an inch in diameter .
The electrolyte was sulphate of copper and dilute sulphuric acid , and was kept agitated during the immersion of the copper rods .
The results are set down in the following Table : TABLE II .
Diameter of Zinc .
Diameter of Copper .
Loss of Zinc in thirty minutes .
inches .
inches .
grs. , 1 30 Y^ 30 2 , 2 2{ 1 30 51 9130 These results perfectly accord with the preceding .
From the consideration of the foregoing experiments , we are led to another important relation of the generating and conducting metals in these cylindrical arrangements , to understand which , it must be borne in mind that the surfaces of cylinders , of equal heights , are directly proportioned to their radii .
Let us therefore imagine an indefinitely small rod of a generating metal placed in the axis of a cylinder of conducting metal of a given diameter , filled with an electrolyte ; upon making contact of the two metals , a current would be established of a definite 141 amount .
The area of the mean section of the electrolyte would be the area of a cylinder placed half way between the cylinder and its axis , or half that of the cylinder ; and it would be the same whether the generating or the conducting metal were the exterior of the two .
Now the amount of the current ought to be the same whatever might be the diameter of the exterior cylinder , for the resistance occasioned by increasing the depth of the electrolyte , that is to say , by increasing the radius of the cylinder , is exactly counterbalanced by the increased conducting power conferred by the increased area of the section of the electrolyte , and vice versed .
The results of the experiments confirm this conclusion ; for upon reference to Table I. it will be seen that cylinders of 1 inch and 23 inches diameter produced , under like circumstances , the same amount of current ; and from Table II .
we learn that cylinders of 24 inches and 5 inches diameter had equal influences .
You may perhaps remember , that in my former communications , from some experiments upon this point* , I had obtained some anomalous results which occasioned me considerable perplexity , but I have since multiplied observations sufficiently to place the confirmation of the law beyond a doubt .
Amongst others , I repeated the experiments with the large battery of ten cells of four inches diameter , in comparisonn with ten of 3 inches , and found the results sensibly the same .
The origin of the error in my former observations I have been unable to detect ; but it is probably to be ascribed to some fault in the connections of the cells .
The advantages of the concentric cylindrical arrangement of the elements of a voltaic circuit are very considerable ; both in the scientific analysis of its complex actions and in its practical applications .
The absolute restriction of the influence of the metallic plates to one side respectively , and the known definite relations of their surfaces and diameters , render the necessary calculations for the former obvious and easy ; while for the latter , the reduction of the size of the generating metal , and the large quantity of the electrolyte which it admits of , give facilities for the maintenance of an energetic and constant current of force which no other arrangement can supply with equal effect .
I have already observedt that the position of the rod within the cylinder is immaterial to the effects , and it is obvious , that wherever placed , their mean distances , and consequently the mean section of the interposed electrolyte , must be the same .
A zinc rod of half an inch in diameter , placed in the axis of a copper cylinder 3inches diameter , produces a certain effect , which is scarcely augmented in an appreciable degree by a second , or even a third similar rod placed in contact with it : the results of experiment were as follows:1 rod = 2'2 2 rods = 2'4 3 rods = 2-5 Each rod separately would have been capable , in its position , of producing the full effect of one ; but each was screened by the two others from the full aspect of the conducting cylinder , and but a slight advantage was gained from the combination by a slight increase of the section of the electrolyte , uncompensated by any increase of distance .
I anticipated that , if each rod were removed as nearly as possible to the sides of the cylinder , so as to be equidistant from the other two , the screening influence could not take place to the same extent , and that a greatly-increased effect would be produced .
A glance at the annexed diagram will explain the difference of the two arrangements ; fig. 1 .
representing a section of the first , and fig. 2 .
of the second .
Fig. 1 .
Fig. 2 .
O)b D OJDb Upon making the experiment , as in fig. 2 , with the rods and cylinder of the last experiment , the result was increased to 3*1 .
I also tested the conclusion with the large battery of ten cells of four inches diameter , and obtained from single rods 10 cubic inches of mixed gases per minute , and from two rods placed as near to the sides of the cylinders as possible , fourteen cubic inches per minute .
The law of the exact compensation of the greater resistance arising from the increased thickness of the electrolyte , by the extension of the area of its mean section , is of course only mathematically correct where the interior wire is infinitely small , but practically the half-inch rods bear so small a proportion to the cylinders which I have been in the habit of employing , that the results are not materially affected by their dimensions .
When , however , the interior cylinders are enlarged , the thickness of the electrolyte is decreased , and the area of its section increased at the same time , and the circulating force rapidly augments .
The results are easily submitted to calculation .
Fig. 3 .
Fig. 4 .
c2 ins .
zc c1 inch zI ''S ' '7B ; 9 t , !
X im mx sIXI ' z an infinitely small zinc rod in the axis .
Let the area of the copper-plate be 100 , the area of the mean section of the electrolyte ( m s ) will be = 50 .
The distance of c from z , or the thickness of the electrolyte , will be 2 inches .
Let the rod z be replaced by a cylinder of zinc z z , two inches in diameter , fig. 4 .
The mean section will be increased to 75 , and the thickness of the electrolyte will be decreased to 1 inch .
The force would , therefore , be increased in the proportion of 50 : 75 for the first , and of 1 : 2 for the second : consequently , compounding the proportions , the force circulating in the first arrangement would be to that in the second as 1 : 3 .
In the preceding observations the cylinders and rods have been taken of equal heights ; when one is shorter than the other , it will be obvious , from a little consideration , that the decrease of length is equivalent to an increase of distance between the two .
Fig. 5 .
, ... .
z. Let cc c ' c represent a section of a copper cylinder , and z'a zinc rod of half its height ; in any action which may take place from the point of the rod z ' to the lower half of the cylinder c ' , the distance between the metals z ' c , or virtual thickness of the electrolyte being greatly augmented , the influence of the lower half will be proportionally diminished .
In some experiments which were carefully made with the calorific galvanometer , an amalgamated zinc rod was successively immersed in the electrolyte in a copper cylinder 31 inches high to the following depths , and with the results set down in the Table : Length of Zinc .
Degrees of galvanometer .
inch 7* inch 19 4 inch 35 1 inch 49 14 inch 67 31 inches 97 In a copper cylinder , twenty-one inches in height , charged with dilute sulphuric acid and sulphate of copper , an amalgamated zinc rod lost 51*5 grains in five minutes ; a rod of half the length lost in the same time 26-1 grains .
In a similar cylinder , six inches in height , charged in a similar manner , a zinc rod of equal length lost 12 144 grains ; a rod of half the length lost 6'6 grains in the same time .
The results , therefore , of each cylinder may be taken as directly proportioned to the lengths of the rods immersed in them .
Let us now turn from the consideration of simple circuits , and examine the law of a series , or of compound circuits .
OHM'S formula for these is no nR+r A in which n represents the number of the series .
Now so long as the external resistance ( r ) interposed in the circuit is merely metallic , the expression accords strictly with the results of experiment ; and by doubling the number of cells at the same time that we double the efficient surface in each cell , we obtain an effect exactly double : thus by the formula 2( no 2nEE 2nE 2 VnR + r ) 2nR n RR + r ; r2 since in doubling the surface in each cell , cceteris paribus , we halve the resistance .
When , however , a voltameter or other chemical resistance is interposed in a circuit , OHM'S formula will not hold , unless the opposite electromotive force which arises from the decomposition of the electrolyte , and consequent accumulation of ions upon the electrodes of the decomposing cell , be taken into consideration .
This is of the same nature as the contrary electromotive force in the cell which we have already pointed out and designated by e ' in the formula E= ( B b e ' ) .
Professor WHEATSTONE , from a series of experiments made conjointly with myself , with my battery , and published in my fifth letter to you , inferred that , if this contrary electromotive force be assumed to be constant and be represented by e and introduced into the formula , thus E--e Rr An tables might be calculated which would represent , approximatively , the quantity of decomposition for any number of cells of a given battery , while the results obtained by regarding the voltameter merely as a resistance , are , it is evident , widely at variance with the truth .
Professor WHEATSTONE devised the following simple means to determine , on this supposition , the values of this contrary electromotive force , and of the added resistance , including that of the voltameter , without having recourse to any other measuring instrument than the voltameter itself .
To obtain the value of the contrary electromotive force , he compared two experiments in which the resistances remained the same , while the sum of the electromotive forces alone varied .
It is obvious that , if there existed no contrary electromotive force , the measured effect in the two cases should be simply as the number of elements in the series employed .
A battery of five single cells should have half the power of a battery of ten double cells ; but instead of this the effects measured by the voltameter were as 6 : 20 .
10Ee 5E -e 10 : 5R + : : 20 : 6 , whence e= 2'857E ... .
( .
) 2 145 The value of r in the formula , i. e. the resistance which the voltarneter and connecting wires add to the circuit , will be ascertained in the following manner .
The comrparison was made with two batteries , one single and the other double , of ten cells each ; the sum of the electromotive forces , therefore , remained the same , while the resistances only varied : 10OE-e loE-e loR+E r : loR : : 12-5 : 20 , whence r= 3333 R ... ( f. ) +r 2 Substituting the values thus found in the general formula n he obtained the nR +r'he obtained the following results : Number of cells ... .
345 10 15 20 Quantity of gas calculated .
336 12s 15^ 17}t cubic inches .
Quantity of gas observed. .
1-1 37 6 12 } 153 17 } cubic inches .
The existence of such a contrary electromotive force , and its great energy , are amply attested by connecting the platinum plates of a voltameter , which has been some time in action , with a galvanometer ; but I purpose to show the general agreement of the amended formula with the results of various and most trying combinations of different batteries , many of which were obtained without the slightest suspicion of the conclusions which might be derived from them .
I say general agreement , for the extremely complicated nature of the actions to be measured , subjected as they were to the necessarily variable influence of circumstances affecting them , the large scale upon which the experiments were carried on , and the roughness and imperfection of the modes of measurement , would necessarily preclude the expectation of absolute accuracy .
The remainder of the experiments , already published in my fifth letter , made with circuits which contained an equal number in series , but in which they were combined as double , treble , quadruple cells , &amp;c. , do not furnish results according with theory so well as might have been expected ; they were therefore repeated with great care , and combined in various ways .
The details of these experiments will presently appear .
The first series was made with a constant battery composed of copper cylinders , six inches high , 3 } inches in diameter , charged in the usual way with dilute sulphuric acid and sulphate of copper .
The first thing to be done was to determine the value of e in these combinations in the modified formula , by comparing the results of two arrangements in which the sums of the-electromotive forces might vary , while the resistance remained the same .
Thus Cubic inches .
5E e in five single cells , 5+ =e 11'25 by experiment ; 10Ee in ten double cells , lo 1 33'7 by experiment ; r therefore 5Ee 10 E e : 11-25 : 337 , or e 249 E. 14-6 To determine the value of r in the formula , we might compare , in a similar manner , as pointed out by Professor WHEATSTONE , the results of two arrangements , in which the electromotive forces being equal , the resistances in the cells alone should vary .
As , however , from the complicated nature of the arrangements , and the variability of different influential circumstances to which I have before alluded , I found it impossible to obtaini two perfectly unexceptionable results for the comparison , I thought it allowable to take the mean of several ; and from this I found that , with a voltameter whose platinum plates are three inches in length by one inch in width , a quarter of an inch apart , and charged with the standard dilute sulphuric acid , ( sp. gr. 1'126 ) , r= 0'541 R in a constant battery of the dimensions just described .
Now if a single cell of such a battery be taken and the circuits closed by a short thick wire , and the zinc rod forming the generating plate of the arrangement be weighed at intervals of five minutes , it will be found to lose 11 26 grs. for every such interval .
This is a measure of the effective force of the circuit ; and its equivalent in mixed gases is 25 cubic inches .
This will be taken as the unit of work in the Table that follows , i. e. ( = i ) , and the calculated results for the different combinations will , in the third and fourth columns , be represented in fractions of this unit .
It is evident , that the amount of zinc , dissolved in such a single circuit , furnishes a measure of the maximum work that any number of such cells , combined in a single E no series , would be capable of performing ; for A , and ni-+ can never be greater than R. , however great the value of n may be , so long as r has a positive value .
In other words , however great the number of cells in a series , it is impossible , so long as any external resistance is interposed , that the result should be greater than that of a single cell in which no exterior iresistance is opposed ; although when r is very small when compared with n R , the results may be virtually equal .
If unity be taken to represent the maximum work that any single circuit can produce , then E will be represented by 1 , and R also by 1 , and E R-1 .
It is evident that in an effective circuit R can never equal E , but for the convenience of calculation it may be assumed to be so ; and as all the quantities in the numerator are compared with E , and all in the denominator with R , the relative proportions will be exact .
Taking the formula nE-e nR+r If E=1 and R= 1 , then e= 2-49 , r= 0'541 .
Substituting different numerical values for n , we obtain for u2 147 single cells , 4 double cells , 4 treble cells , Calculation .
4 2-49 1*51 Cubic inches .
4+ 0541 4-541 = 03325 = 831 4 -2-49 151 4 2541 2 =-0'5942 = 1485 4+0 2541 -~q0'541 4 2-49 1'51 -4 1'871 = 0'8071 = 20'17.-+ 0-541 4 2'49 1P51 4 quadruple cells , 4= 141 + 0-541 09799 = 24'5 4 2'49 1'51 4 quintuple cells , 4= 1'126 = 28,15 -+ 0'541 5 -249 2-51 5 single cells , 5 -0.541 5.541 '453 = 11'33 5+ 0,541 5-541- ?
~ ?
3 Experiment .
Cubic inches .
7-5 13-7 21 25'5 30 11*25 5 double cells , 5 treble cells , 5 2-49 2-51 5+ 541 3041 -+ 0541 5 2-49 2*51 5 2'208 + 0o541 08254 20'63 = -1137 =28-42,.~5 --q 5 249 2'57 5 quadruple cells , 2-49 15791 T+ 0-541 10 single cells , 10 double cells , 1401 = 35-04 10 249 7-51 10 + 0'541 10-541 014 1781 10 2-49 7'51 10 ll.3 1'355 = 33'88 10 5 5-541 --0'541 15 249 12-51 15 single cells , 15 + 0541 15541 0-8117 = 20'29 18-7 20 2,49 17*51 20 single cells , 120+2549 17254 22 . ? ?
20 + 0541 2054108524 2 -3 22 .
The agreement of the calculated and experimental results under such complicated circumstances , as shown in the last two columns of the preceding Table , must , I think , be deemed very satisfactory ; and it is worthy of remark , that the result just named , of the independent experiment with the single cell , 25 cubic inches , is altnost identical with that deduced from the experimental determination of five cells ; taking 11-25 cubic inches to represent accurately the fraction 0'453 : and indeed agree very 20-5 28-7 35'2 15-7 33-7 148 closely with the calculated results of the above Table , whatever combination be taken as the foundation of the calculation .
In the experiments already alluded to , which I performed in conjunction with Professor WHEATSTONE , the cells of the battery used were of the same character as the last , but with an efficient length of 20 inches , or 3'33 times greater .
The duration of each experiment was in this case one minute .
Upon making the calculations for these , E being = 1 , R= 1 , e was found =2 85 , by comparing the results of five single with those of ten double cells ; and r ( by a mean of seven experiments ) = 1'757 .
Hence we find , 5 single cells , 5 double cells , 5 2-85 2'13 5+ 1757 6757 5 285 2*13 5 757 4257 2 Calculation .
Cubic inches .
= 0-3182 = 7'31 = -0505 =1 159 5 treble cells , 52'85 2*13 + --l , --0'6281 = 14'4 + 1'757 -T 5 2-85 2'13 5 quadruple cells , 5= 307 4+ 1'757 = 0'715 = 16'44 10 single cells , 10 757= 1757 = 0'6081 = 1398 o1-2n 7 757 10 double cells , -0 +7= 1058 = 2433 + 1-757 52 14 15'66 12.25 20 20 2'85 17'13 20 single cells , 20 + l'77 21757 = 0-7882 = 18'55 17'25 .
These results again exhibit a general accordance with the calculation , but by no means so close as the preceding .
I was therefore induced to repeat the experiments with great care .
The following Table shows the results , which it will be seen closely correspond with those deduced by the formula .
By a mean of fourteen experiments r was again determined to be in this battery 1-725 R. Taking , as before , E= 1 , R= 1 , and e= 249 , we obtain from 5 single cells , 5 double cells , Calculation .
~5 --~~ 2'49 2 ' &amp; Cubic inches .
; ) 2-49 2'51 5+ 725 625 -= 0'3732 = 858 5+ 1=25 ; 67255 2'49 2'51 5 4'225 = 05941 = 13'66 + 1P725 2 Experiment .
Cubic inches .
6 11 Experiment .
Cubic inches .
8-875 13.5 149 Calculation .
Experiment .
5 2Q49 2o51 Cubic inches .
Cubic inches .
5 treble cells , = 0-738 1697 17'0 1-'725 5 2'49 2'51 5 quadruple cells , 97 437 = 19 20'0 10 single cells , 10 +1 725 = 1125 = 06408 = 1473 15'25 10 2'49 7'51 10 double cells , =1 116 = 2568 25'5 + 1725 ~ 6-7202 202'49 17'51 20 single cells , 20 + 1725 = 21725 08062 = 18'54 18'00 .
The maximum work of a single circuit of this battery was found to be 9*95 grs. of zinc per minute , which is equivalent to twenty-three cubic inches of the mixed gases .
A similar agreement of this independent result which has been taken as the unit in the preceding table ( for = 1 ) with those which would be afforded by any combination of cells taken as the foundation of the calculation may be also observed , as in the case of the table deduced from experiments with the smaller battery .
When a number of cells of different power are included in the same circuit , the expression becomes no + n1 Ee ( n + n ' ) R+rA supposing that R remains the same as in the regular circuit , and E ' represents the electromotive force of the new element , and n ' the number of the new elements included .
It will further , on a little consideration , be obvious why a half-zinc rod may be substituted for a whole one in a series , without any perceptible diminution of the effect , as I found upon a former occasion* .
The effect of diminishing the length of the rod is principally to increase the distances between the metals , as the dimensions of the mean section of the electrolyte will scarcely be altered , owing to the comparatively small surface of the generating metal , even when entire .
The general formula will then become no e n~--eA ~R+ nR R~ +r = A , when n ' represents the number of ordinary zinc rods , n " the number of shortened ones , and R ' the increased resistance offered by each of the latter .
The mean distance between the metals will , perhaps , be increased one-third by halving the length of the zinc ; and the resistance R presented by that cell , varying directly as the distance , will be 14 instead of one .
The formula , therefore , for 20 small cells with 19 zinc rods of the ordinary size , and 1 short one , becomes 20:2-49 17-51 20'97 cubic inches 19 + 1-33 + 0-541 = 20871 2097 cubc ches instead of 20 2.49 17-51 20 + 0541 = 2054121'31 cubic inches , 20 + 0:541 20'541 when the zincs are all of the full length .
Thus with a small series of five cells with entire rods 5 -2.49 2-51 5+ 0'541 5'541 = 11133 cubic inches ; with four entire rods and one half rod , 5-2-49 2-51 = 10'69 cubic inches ; 4+ 1333 + 0o541 5871 = 069 cubc ches the differences not being appreciable in the usual mode of measurement .
In an arrangement containing one or more reversed cells , the formula becomes ( n n ) Ee ( n + nt)R +r where n ' represents the number of reversed cells .
My fifth letter contains the results of experiments with the battery under such circumstances .
Upon comparing these results with those of the formula just given , the discrepancies were found to be constant and considerable ; such however as might be accounted for by supposing that each reversed cell introduced , in addition to the reversed current , an extra resistance .
Upon searching for this resistance it was proved to exist , and its cause manifested by the following experiments .
Ten small cells , charged in the usual manner , were arranged in series , including a voltameter , and an additional cell with a similar charge ; substituting a copper rod in the interior for one of zinc .
In four minutes seven cubic inches of gas were collected in the voltameter ; the action then suddenly declined in intensity , and in the next four minutes only 34 cubic inches of gas were given off .
Upon examining the copper rod it was found coated with a film of oxide : this was wiped off , and a bright metallic surface was again exposed , and on once more connecting the battery , seven cubic inches of gas were given off , as before , in four minutes .
The quantity again declined .
This was several times repeated , and always with the same general results .
It is hence evident , that when copper is made the zincode in the series , a layer of oxide is deposited upon it , which is not immediately dissolved by the acid , and offers a resistance which will vary according to its thickness ; and this again will much depend upon the size of the surface .
151 This is , moreover , not the only point for consideration ; for it is probable that the hydrogen accumulated upon the zinc of the reversed cells would exalt its electronotive force , so that E would be somewhat increased ; and upon milaking the calculation upon these amended data the formula becomes no n ' even -n ' E ' e --. .
t -(n + n')R+r + nr^ A or nR+ R +n r ' taking in the first formula r ' to represent the additional resistance introduced by each reversed cell , and in the second R ' as the total resistance in each reversed cell , and E ' the increased electromotive force in each of the same reversed cells .
In the following calculation E= 1 , R= 1 , e= 2'85 , r= 1'725 , E ' = 11 , r1 = 05 , .
It is assumed that r ' is a constant quantity , which may be pretty accurately true when the copper surface is so large in relation to the zinc as in the present case .
Calculation .
Experiment , 20 285 17 Cuic inches Cubic inches .
20 direct , 20 1'5 25 1818 175 20 + 1'725 21 725 19285 1'1 15-05 1 reversed , 20 + 1-725 + 0-5 22.225 15 57 15'5 18 2-85 2-2 12'95 2 reversed , 13,1 12-75 20 + 1'725 + 1.0 -22'275 1 17 2-85 33 1085 74 13 reversed , 20 + 1'725 +1.5 = 23.225 16 2-85 44 875 4 revesed 20 + 1725 + 2-0 23.275 8'48 8'5 15 2'85 5'5 6*65 5 revesed , 20 + 1'725 + 2-5 24'225 6'31 5.5 14 2-85 6-6 4-55 6 revesed , 20 + 1-725 + 3'0 24'725 23 3 13 2-85 7-7 2'45 7 reversed , 20 + 1.725 + 3-5 2522 2'23 1'625 12 2.85 8-8 0-35 8 reversed , 20 + 1.725 + 4o 25.725 0'31 16 The agreement of the experiments with the calculations is not as close as before , especially in the lower part of the table , but may I think be deemed satisfactory , as a first approximation to the solution of a problem of a most complicated nature .
The influence of the dimensions of the plates of a voltameter upon the amount of decomposition may also be submitted to calculation in the same way .
This influence will , of course , be most perceptible when a small number of elements presenting a large surface is employed ; whereas , when a numerous series is made use of , the dimensions of the electrodes are of little consequence .
Some experiments which I have made with a large voltameter , kindly lent to me for the purpose by Mr. GASSIOT , will place this in a striking point of view .
The voltameter consisted of five pairs'of platinum plates , each four inches by 3 ?
inches , at an average distance of half an inch apart .
These were so arranged that any number of them might at pleasure be connected with a battery .
20 cells of the small battery were so arranged as to form a series of 5 quadruple cells , and then connected with one pair of plates of the voltameter .
By a mean of two experiments they gave 26'2 cubic inches of gases for five minutes .
When all the plates of the voltameter were connected with the battery , the product of gases for five minutes was 32 cubic inches .
The same battery arranged to form a series of 20 single cells , furnished with one pair of plates 16 cubic inches , and with all the plates the result was the same .
Now by experiment , if E= 1 , e= 2449 , R= 1 , r= resistance with one pair .
20 2*49 _1751 ( 1 .
) 20 with one pair = 20+ +r 27+= 16 cubic inches .
----2o+r r 2o0-r 20 2'49 17-51 ( 2 .
) 20 with five pairs 49 7= 16 cubic inches .
t~L ~ 20 +r 20 +r ~~5 5 ~~~~r ~5 2*49 2'51 ( 3 .
) 5 quadruple with one pair =i= 26-2 cubic inches .
- ... 5 : !
.2 + -26'2 cubic inches .
+r 45 2'49 2'51 ( 4 .
) 5 quadruple with five pairs =r =.2 .
=32 cubic inches .
_4+ / 5 1-25 +r455 Since the electromotive forces in the two last expressions are the same , we can , by comparing them , ascertain the value of r. Thus 1.25 + r : 1'25 + : 32 : 262 18*125 1 r-5254 =3 nearly .
524 3ne Now substituting this value of r in the expressions ( 1 .
) and ( 2 .
) , and adopting the experimental result of 16 cubic inches , we obtain for Calculation .
17'51 ( 1 .
) the fraction 2033 = 16'2 cubic inches .
( 2 .
) the fraction -6 -0= 16-0 cubic inches .
The calculated results it will be seen almost coincide with each other , as do the experiments .
By the substitution of different values for R and r in the formula , it will be found that every different arrangement must have a distinct number in series , which it will be most advantageous to work with , and this number will vary in the same arrangement , with the nature of the electrolyte , and also with the size of the battery plates .
It will appear from calculation , that the most advantageous combination is that in which the value of A ( in the formula -n E+ A ) most nearly approaches to 0'5 .
It will therefore vary even in batteries of the same chemical construction ; increasing as R diminishes in proportion to r : or in other words , when the plates are large , a more numerous series is required , than when small , to produce the most advantageous results .
This is likewise the case when the exterior resistance is increased : in both cases R is virtually diminished in respect to r. It is evident from the preceding observations , that all the comparisonts hitherto made by different experimenters between the general relative powers of different batteries are faulty , inasmuch as they only hold good in the particular cases to which the experiments are limited ; and one battery of a certain size may be preferable for one kind of decomposition , and yet may allow of considerable useless expenditure of power when a different electrolyte is subjected to its action .
As , however , great stress has , by some , been laid on comparisons of this kind , it may not be amiss to give a few experimental results .
In the following arrangements the cylindrical form was employed .
In each case the platinum cylinders were 4 inches high , and 1 inch in diameter .
The zinc rods were half an inch thick , the exciting fluid was placed in a porous earthenware tube 1 inch in diameter .
Three cells was the number employed in each experiment .
The measures employed was the quantity of mixed gas produced from the voltameter during 5 minutes .
Exterior liquids .
Exciting liquids .
Gas in five minutes .
cubic inches .
Acid sulphate of copper ... ... ... ... ... ... Dilute sulphuric acid , specific gravity 1*126 3-0 Nitric acid , specific gravity 1'40 ... ... ... .
Dilute sulphuric acid , specific gravity 1'126 14'0 Bichromate of potash , specific gravity 1*050. .
Dilute sulphuric acid , specific gravity 1'126 3'1 Bichromate of potash , sulphuric acid ... ... .Dilute sulphuric acid , specific gravity 1'126 5-5 Nitrate of copper , neutral saturated solution Dilute sulphuric acid , specific gravity 1'126 3-5 Nitric acid , specific gravity 1'40 ... ... ... .
Dilute nitric acid , specific gravity. .
1 056 8'5 The eligibility of a liquid in the construction of a battery will , of course , be much influenced by its conducting power : and it would at first sight appear that this might be easily determined by placing the different liquids , in succession , in the same voltameter , or experimental cell , and transmitting through them a constant current of known power ; measuring the retarding influence of each by another voltameter charged in the usual way with dilute sulphuric acid , and included in the same circuit .
The following are the results of some experiments so performed .
The current from ten cells of the small constant battery was employed : Liquid tested .
Gas from voltameter in five minutes .
Cubic inches .
Nitric acid , specific gravity 1*4 ... ... 11 Nitric acid +8 sulphuric acid ... . .
11 Dilute sulphuric acid ( standard ) ... ... 9'2 Dilute sulphuric acid saturated with sulphate of copper .
9'2 Saturated solution of sulphate of copper ... ... .
79 Bichromate of potassa , specific gravity 1-050. .
5-6 154 Let us however consider these results with reference to the formula .
The expression for the arrangement used becomes nEee nTR +r+ ri ' in which e ' signifies the contrary electromotive force , introduced by the accumulation of the ions on the plates of the voltameter containing the liquid tested , and r ' the resistance offered by the same .
It is clear that we cannot estimate r ' , which we are attempting , unless e ' is known or remains constant : now e ' is not constant , since with nitric acid it vanishes probably altogether , and varies with each of the other substances employed .
If chloride of platinum were not too expensive to allow of its being employed as the exterior part of the electrolyte in contact with a platinum , conducting plate , e ' , or the contrary electromotive force would be wholly annihilated , as nothing but platinum would be thrown down upon the platinum , and it would constitute the most perfect possible arrangement , but would not much , if anything , exceed the efficiency of nitric acid .
In the nitric acid battery the electromotive force is nearly double that of the sulphate of copper arrangement , and consequently one cell of that construction is capable of effecting the decomposition of dilute sulphuric acid .
It is evident that a similar series of calculations might be made for this battery and others in which different electrolytes are employed ; R varying as before with the size and distance of the plates .
I have , however , done enough for the accomplishment of my present purpose ; but must not conclude without expressing my obligations to my friend Dr. MILLER for his able assistance , both in the performance of the experiments and the calculations of the results which I have now the pleasure of communicating to you .
I remain , my dear FARADAY , Very faithfully yours , J. F. DANIELL .
King 's College , 12th April , 1842 .
x2 155

