ELECTH~XOTIVE FORCE OF ALLOYS IN A VOLTAIC CELL. 1031 LXXXIII.-77w Eloct?.omotiz:e E’orcc of Alloys in CL Voltaic Cell. By A. P. LAURIE, &LA. 0s former occasions, I have brought before the Society (Trans., 1889, 65,677) the results obtained in testing the E.M.F. of alloys in voltaic cells with the view of throwing further light on their constitution. Briefly, the method can best be explained by taking an example. If we place a plate of copper and a plate of zinc in a solution, we shall get an EMF. between the zinc and copper which can be mea- sured on a Thomsoii electrometer; if, on the other hand, we sub- stitute a copper plate for the zinc there will be practicallyno E.M.F. between copper and copper. If we now attach a small piece of zinc to the surface of one of the copper plates beneath the liquid we shall again get approximately the E.M.F. between zinc and copper, and if we suppose that the zinc, even if present in minute quantities, was mixed with the copper throughout its mass, itt wouid polarise the copper plate, and the compound plate would give an E.M.F.approximately the snme if it consisted of zinc alone. If, however, the zinc had entered into combination with the copper, i t would no longer be able to polarise the copper plate, and we should be dealing with a new metal instead of ft mixture of two metals. Furthermore, if zinc forms such a compound with copper, it may be expected to have a definite composition, and consequently, if the proportion of zinc present in this plate is gradually increased, we should expect to reach a point at which the copper was saturated with zinc, and any additional zinc being merely mixed with the metallic compound would cause a r abrupt rise in the E.M.F.I have shown i n former papers that such an abrupt rise takes place in the case of the zinc-copper alloys, the discontinuity in the curve representing the E.M.F. occurring at a point corresponding to an alloy of the formula CuZn,. I n the case of the copper-tin allojs, the discontinuity occurs at Cu3Sn. On comparing this result with the measurement of the1032 LAURIE : THE ELECTROMOTIVE FORCE electrical resistances of the copper-tin alloys by Professor Lodge and Professor Roberts-Ansten, a confirmation of the view that Cu,Sn is a compound is obtained. As is well known, Matthiessen obtained four types of curves as the result of his measurements of the electrical resistance of alloys.First he found that in certain alloys the conductivity was propor- xiooal to the amount of each metal present, the curve being a straight line joining the conductivity of the two metals. In the second t-pe of curve, the conductivity rapidly diminislled on the addition of only a small quantity of the metal having the lower conductivity, and then became practically a straight line join- ing the cur\-e representing the metal of lower Conductivity. I n the third type of curve, the electrical conductiritg diminished (rapidly on the addition of a smsll quantity of either metal to the other, thus producing a U-shaped curve. In all these cases there is, as Matthiessen truly points out, no indication of the formation of a compound, the changes in conductivity being brought about by such small traces of the metals that we must regard it rather as a case of allotropic modification.There was, however, another form of curve which he obtained ill the case of the gold-tin alloys in which discontinuities occurred in the case of intermediate mixtures of the metals, the curve having the form of a W, and he suggested that the two minimum points and the maximum point represented three different compounds between gold and tin. It is, howevel-, I think, equally open to 11s to regard these results as Seing due t o the forma- -tion of a single compound of the percentage composition correspond- ing to the maximum point of the curve, and to suppose that the two portions of the curve on each side of the maximum point are similar ‘to tile U curves obtained in other cases, the compound behaving like a new metal.In the case of the copper-tin alloys referred to above, Lodge aBd Roberts-Austen hare found a maximum point 4corresponding to the alloy Cu3Sn. In the case of the gold-tin alloys, I have already shown that a, discontinuity occurs in the curve of E.1I.F. at the percentage com- position AuSn which corresponds with the maximum point in Xatthiessen’s curve of electric conductivity, and that in the case of the tin-lead, tin-cadmium, zinc-lead, tin-zinc, and lead-cadmium a11oys which Matthiessen regards merely as mixtures, there is no indication of the existence of a compound, the E.M.F. rising in each case on the acidition of a small quantity of the more positive metal, As, therefore, the resuits obtained by this new method have so far agreed with those obtained by measuring the electrical conductivity of the alloys, I thought it as well, before using the method for ex- ploring new regions, to apply i t systematically to each of the groupsOF ALLOYS IN A VOLTAIC CELL.1033 of alloys which are given in Matthiessen’s paper on the electric con- ductivity of alloys (Trans. Roy. SOC., 1860, vol. 150, p. 161). The number of groups of alloys of which he determined the electrical conductivity in this paper is 19; I have already published the re- sults of my examination of the E.N.F. of six of these, leaving 13, and now add a considerable number of fresh results. These can be divided into tmo groups, namely, those which, on the introduction of 2 per cent.to 3 per cent. of the more positive metal, give a t once the full E.M.F. due to that metal, and those which show a gradually increasing E.M.F. as more and more of the more positive metal is added, approximating to a maximnrn on the addition of some 20-30 per cent. In no case has a discontinuity heen discovered such as that obtained in the case of the gold-tin, copper-tin, and copper- zinc alloys, thus confirming by my method Matthiessen’s conclusion that in the case of these alloys, with the exception of gold-tin, nothing in the way of chemical corcbination takes place. These results, together Iyith those already published, include 16 of the 19 alloys described by Matthiessen, the silver-tin and gold-copper allojs being excluded, as, owing to some cause which I have not yet been able to discorer, the readings obtained with these alloys and metals were so very irregular that I do not feel justified in publish- ing them until I hare discovered the source of these irregularities ; there was no indication, however, of the existence of a compound in the case of these alloys.With reference t o the two groups of alloys, namely, those giving an immediate high E.M.F., and those which give a gradually rising one, there is no sharp dividing line, as it is rather a difference of degree than difference of kind ; nor do the two groups correspond exactly with Mat thiessen’s two groups, namely, alloys which are merely mix- tures, and alloys which produce an allotropic change.It is furthermcre to be noted that whilst careful repetitions of the measurements a1wa-j-s result in a ainiilar curve being obtained, the readings will not be exactly the same in each case for an alloy of the same percentage composition. This is, I think, to be expected if we consicler these results as due to something corresponding to the case of an aqueous solution of a salt. It is easily understood that on the solidification of such a solution, the conditions will not be precise]F the same on any two separate occasions, or in t-n-o different parts of the alloy, and consequently the E.N.F. obtained must differ slightly for different preparations of the alloy. The method employed has been to prepare small samples of the alloys from approximately pure metals, and, by means of a Thornson quadrant electrometer, to test the EMF.in voltaic cells of various construction against that of a standard Daniel]. I shall first give the TOL. LXV. 4 D1032 LAURIE : THE ELECTROMOTIVE FORCE results obtained with those alloys in which the full E.M.F. is given as soon as a small quantity of the more positive metal has been added. Bismuth and Tin. The cell used consisted of a solution of stannous chloride around the tin plate, and of common salt around the bismuth plate, the two solutions being sqmrated by a plug of filter paper in a narrow glass tube. Under these conditions there was a slight E.M F. of 0.07 volt between bismuth and bismuth, but this does not affect the results. The E.M.F.s obtained were as follows.The results given are the means of three or four readings, the alloys being cleaned and scraped between each observation. Tin is positive to bismuth. Tin against bismuth .......................... 0.119 volt. 0.119 Alloy (tin 5 p. c., bismuth 95 p. c.) against bismuth ,, ,, (tin 50 ,, bismuth 50 ,, ) ,, 7 , 0.119 ,, Hisnzuth and Lead. In this case, the bismuth was i n common salt, and the lead in :e paste of lead chloride. Lead is positive to bismuth. Lead against bismuth ......................... 0.139 1-olt. 0.134 Alloy (lead 5 p. c., bismuth 95 p. c.) against bismuth ,, Bismitth a d Ziuc. In tbis case, both metals were immersed i n a single solution of Zinc against bismuth.. ....................... Alloy (5 p- c. zinc, 95 p. c. bismuth) against bismuth This result is slightly lower, showing that there is nc sharp line coinmon salt.@iY2 volt. 0.758 ~, between the allojs in this group and in the next. Bisnzzith a d Gold. These alloys were tested in a solution of common salt. Alloy (gold 95 p. c., bismuth 5 p. c.) against gold. Bismuth against gold.. ....................... 0.41.7 volt. 0-412 ,, Bismuth and Silver. Bismuth in common salt; silver siirrounded by a paste of silI-er Silver against silver.. ........................ Bismuth against silver.. ...................... 0.102 ,, Alloy(974 p. c. silver, 2$ p. c. bismuth)against silrer ,, chloride. Bismuth is positive to silver. 0.036 volt. 0.112OF -4LLOF'S IK A VOLTAIC CELL. 1035 Gold and Silrer. Both in a solution of common salt. Silver is positive to gold.Silver against gold ........................... Alloy (95 p. c. gold, 5 p. c. silver) against gold. .. 0-117 volt. 0.219 ,, Cadmium and Zinc. Both in a solution of common salt. Zinc against cadmium.. ....................... ,411oy ( 5 p. c. zinc, 9.5 p. c. cadmium) against cadmium Zinc is positive to cadmium. 0.239 volt. 0.216 ,, Antimony and Tin. These metals were tested in a solution of common salt, Tin is positive to antimony. Tin against antimony.. ....................... Alloy (5 p. c. tin, 95 p. c. antimony) against antimonr To this list must be added the former determinations of tin-lead, zinc-lead, zinc-tin, and tin-cadmium. We next come to the alloys which shorn a gradual rise of E.M.F. as the more positive metal increases in amount. The first examined were the lead-antimony alloys ; several different samples of these were prepared and tested, using different types ot \-oltaic cells.0.211 Folt. 0.188 ,, Antinzony and Lead. The readings were taken with both the alloy and the antimony in a cell of sodium chloride. E.M.F. I. Sntimong containing 4.76 per cent. lead-antimony 0.123 volt. 7, Y 7 9-05 ,, 7 :, 0-169 ,, 7 7 7, 16% ), 7 ., 0.215 ,, 7 7 7 13.4 ,, ., 0.182 ,, Antimony against lead. ....... .-. 0.246 ,, The readings given in I1 were taken with a fresh set of alloys, also in a single cell of sodium chloride. E.M.F. 11. Antimony containing 4.74 per cent. lead-antimony 0.057 volt. 7 9 7 : 9.209 ,, Y 7 3 9 0.112 ,, 7, 7 9 13.03 ,, 2 ) 7 7 0.1349 ? 7 7 , 7 7 17.03 ,, , 77 0.155 7 7 Antimony against lead ..........0.207 ,,1036 LAURIE : THE ELECTROMOTIVE FORCE The set of readings, 111, taken with the same set of alloys as in 11, only with the antimony in an inner cell containing antimony chloride, and the alloy in an outer cell containing sodium chloride, gave the following results. 111.-Antimony against antimony. .................. E.M.F. 0.285 rolt. Antimony containing 4-74 per cent. lead-antimony 0.358 7, 77 7 9.209 ,, , .) 0.415 ,7 7, ), 13.03 ,) ., 0.434 ), 7 9 7 7 1i.03 ,, 9 .) 0.475 ,) Antimony against lead.. ...... 0-494 7 , The set of readings, IV, taken with the same alloys undsr the same conditions as before, except that the alloys were in an outer cell of potassium nitrate, gave the following results. 1V.-Antimony against antimony.. .................E .M .F . 0.259 volt. Antimony containing 41-74 per cent. lead-antimony 0.342 ,7 9 7 $ 7 9.209 1, . 7 7, 0.386 9 7 9 7 ,, 13-03 , 7 7 7 ), 0.413 ,. 9 9 7 7 17-03 ,, ,? .? 0-415 ), ,4ntimony against lead ........ 0-456 ,) I n V, we have the same set of alloys read again with the alloy in an outer cell of magnesium sulphate. V.-Antimony against antimony.. .................. E.M.F. 0.254 volt. Antimony containing 4-74 per cent. lead-antimony 0-2903 ,, 7 7 7 7 9-209 ., .) .) 0 342 ), 7 , 13.031 7 7 ,, ,, 0.386 :, 7 7 7 17.03 7 7 ., ,) 0.435 ,, Antimony against lead. ......... 0.451 . 7 It will be noticed that in each of these tables the E.3l.F. rises rapidly in a continuous curve. The following are further examples of a gradually rising E.1I.F. Lead aid Gold.These alloys were tested in a solution of common salt. Lead against gold ............................ Lead is positive to gold. 0.523 Tolt. Alloy ( 5 p. c . lead, 95 p. c. gold) against gold .... 0-353 .. ), (7 p. c. lead, 93 p. c. gold) 7 7 .... 0-461: .. ,) (10 p. c. lend, 90 p. c. gold) ,, .... 0.317 ..OF -4LLOTS IN -4 VOLTAIC CELL. 1037 Lead and Silver. These met(a1s were tested in common salt. Lead is positive to silver. Lead against silver.. .......................... 0.505 volt. 0.406 Alloy (lead 5 p. c., silver 95 p. c.) against silver. .. >, ,, (lead 15 p. c., silver 85 p. c.) . . . . . 0.486 ,, In the former paper (Eoc. cit.), the E.M.F. obtained between cad- mium and lead is given as being 0.322 rolt, whilst an alloy with 3 per cent. cadmium gave 0.264.This seems to show that this alloy belongs to the present group, and I hare, therefore, tested its E.M.F. again in a solution of common salt. ,, (lead 10 p. c., silver 90 p. c.) . . . . . 0.441 ,, Cadmium and Lead. Cadmium is positive to lead. Cadmium against lead. ........................ Alloy (5 p. c. cadmium, 95 p. c. lead) against'lead . . 0.197 volt. 0.176 ,, t h u s showing a slight increase of E.M.F., which, however, disappears on amalgamating the alloy, so that this alloy can probably be placed in the first group. The copper-silver alloys I have not experimented on, and, as already stated, satisfactory results were not obtained with gold-copper and silver-tin. These results, including 15 of Xatthiessen's alloys and one other alloy, namely, bismuth-zinc, confirm in each case his conclusion that, in these alloys no compound of the metals exists.I have also repeated the former measurements of the tin-gold alloys, nsing a solution of common salt in place of the solutions formerly used. The results indicate the existence of a gradually rising E.M.F., suggesting some heat of solution, as well as the abrupt rise of E.31.F. which marks the passage over the compound. Gold and Tiw. Tin against gold.. ............................ Alloy (10 p. c. tin, 90 p. c. gold) against gold ..... ,, >, 26 7 7 :7 '28 Y , 7 , :34 7 7 ,, 36 ,, .................. ., 40 ,, .................. >, 50 ?, (23 p. c . tin) against gold.. ................ 9 7 .................. 3 - .................. 17 .................. 7 , 7, 7 7 ..................0.435 volt. 0.191 ,, 0.261 :, 0.266 ,, 0.311 ,, 0-297 ,, 0.311 ,, 0-396 ,, 0.428 ,,1038 LAURIE : THE ELECTRONOTlVE FORCE thus indicating a gradual rise of E.M.P. up to the 36 per cent. alloy when a sudden rise of nearly one-t,enth of a volt takes place on passing this point. The results of amalgamating these gold-tin alloys is somewhat curious. In both the copper-tin and the copper-zinc alloys the com- pound was not affected by amalga,mat;on ; but in the case of tin-gold the amalgamation breaks up the compound, the result being a gradual rise of E.M.F. throughout the series of alloys. Tin aiLd Gold Amalgamated. Tin against gold .............................. Alloy ( 5 p. c. tin, 9.5 p. c. gold) against gold ..... .. 10 3, qainstgold.................. 7 ) 23 .) .................. ,l 26 - 7 7 7 .................. ,? 34 3 , .................. >, 36 9 7 77 .................. ,, 40 ., .................. 7 7 5 0 :, >. .................. ?, 7 7 0.492 volt. 0.270 9 ) 0.290 .) 0.280 ., 0.310 ., 0.320 7 , 0.s40 ,. 0.330 .: 0-360 ., This curve has not been investigated beyond this point. This result suggested the possibility of dispiacing one metal in the compound by another, and I accordingly decided to try and displace the tin in the copper-tin alloy by zinc. The results obtained cer- tainly indicate that t h i s takes place. The copper-tin alloy contained 35 per cent. of tin, which is a little less than tbat necessary to form a compound between copper and tin, and pave an E.M.F. against copper of 0.057 volt.To this the same weight of zinc as there was vopper in the alloy was added, and the triple alloy gave an E.M.F. against copper of 0.259 volt, whilst tin against copper gave 0.272 volt, and zinc against copper gave 0.840 volt, and an alloy of zinc 50 per cent. with copper 53 per cent. against copper gave 0.098. Con- sequently the E.M.F. obtained from the triple alloy agreed closely with that obtained from tin, but was much below that obtained from zinc, indicating that the zinc had combined with the copper and turned out the tin. If the results obtained by Matthiessen and others as to the existelice of compounds between metals are examined, it will at once be evident that there is no apparent explanation of why certain rnehds do, and other do not, combine, nor do the results bear any obvious relationship to Mendelkeff’s table. It is remark- able, for instance, that two metals like lead and zinc do not combine, and it occurred to me as a possible hypothesis that iheseOF ALLOTS IS A VOLTAIC CELL. 1039 results might be due to the dissociation of the compound at the tern- perature of fusion of the metals. I only mention this possible expla- nation to point out that I have been unable to obtain any evidence of its correctness. I find that alloys of lead and zinc which have been lowered iii melting point by the introduction of bismuth or mercury, still indicate the presence of free zinc, showing that lowering the melting point does not enable the zinc and lead to combine; the explanation, therefore, must be sought elsewhere. The inyestigation of this question seems to me important, as we do not know a t present how many of the elements will combine two and two, aucl investigation of a sufficient rumber of alloys should finally lead to some conclusions from which a law of combination among the elements might be derived.