II .
" Analytical and Synthetical Attemnpts to ascertain the cause of the differences of Electric Conductivity discovered in WVires of nearly pure Copper .
" By Professor WILLIAM THOMISON , F.R.S. Received December 22 , 1859 .
Five specimens of copper wire No. 22 gauge , out of a large number which had been put into my hands by the Gutta Percha Company to be tested for electric conductivity , were chosen as having their conductivities in proportion to the following widely different numbers , 42 , 71U3 , 84 7 , 8654 , and 102 ; and were subjected to a most careful chemical analysis by Professor ILofniann , who at my request kindly undertook and carried out what proved to be a most troublesome inivestig , ation .
The following report contains a statement of the results at which he arrived : 'Royal College of Chemistry , March 10th , 1858 .
" SJR,1 10ow beg to communicate to you the results obtained in the anialysis of the several varieties of copper wire intenided for the use of the Transatlantic Telegraph Company , wh-ich you forwarded to me for examination .
" I have limited the inquiry to a minute qualitative analysis of the wires , to a very accurate determination of the amount of copper , and an approximative determination of the amount of oxygen .
The qualitative analysis has been repeated several times with as considerable quantities as the amount of material at my disposal permitted .
The quantitative determinations of the copper have been made with particular care , and after a lenigthened scrupulous inquiry into the limit of accuracy of which the method employed is capable , I am convinced that the true per-centages of copper cannot be more than 0 ) I per cenit .
either above or below the means of the determinationls , the details of which I give you in the Appelndix .
6The following , Table containis the results fuirnished by analysis Conductivity of the l wire , in relative 42 .
71-3 .
847 .
8604 .
102 .
measure* .
( Copper .
|Copper .
Copper .
Copper .
Copper .
Iron . !
Iron .
Iron .
Iron .
Iron .
Qualitative analysis ijNickel .
Nickel .
Nickel ( doubtful).-Nickel ( doubtful ) .
Oxygen .
Arsenic .
Oxygen .
Oxygen .
Oxygen .
( Oxygen .
Per-centage of copper 98r76 9920 99r53 99.57 99.90 Amount of impurities .
1-24 0-80 0 47 0 43 0 10 100,00 10000 10000 10000 10000 " Since it appeared probable that the extraordinary difference in the conductivity of the several specimens was due rather to non-metallic impurities than to metallic admixtuires , careful experiments were made in every case for the detection of sulphur .
In none of the specimens was it possible to discover the slightest trace of sulphur .
Qualitative experiments having established on the other hand the presence of oxygen , probably in the form of suboxide of copper in every one of the specimens , an attempt was made to ascertain the quantities by determining the loss which the wire after rolling suffered when heated in an atmosphere of hydrogen , and by simultaneously estimating the quantity of water formed .
" In this experiment , the details of which are given in its Appendix , the following numbers were obtained Conductivity ... ... ... ... .
42 71.3 847 1 864 102 Percentage of Oxygen ... ... 0-087 0-119 0-172 0-159 0-193 " Unfortunately the same reliance cannot be placed upon these numbers as upon the preceding ones , since the method employed involves many sources of error , and want of material precluded the possibility of repeating the experiments .
" From the preceding antalysis , it is obvious that the amount of impurities in the several specimens examined is small , varying as it does between 0 10 and 124 per cent. The number of foreian constituents also is comparatively small .
I should , however , state that the analytical results which I have given do not exclude the presence of exceedingly minlute quantities , eveni of other metals which might have beeil detected if larger quantities of copper could have been sul)mitted to analysis .
Some years ago , AMax Duke of Leuchtenberg* examined the black precipitate formed at the aniode in the electrotype process , during the decomposition of sulphate of copper by the galvanic current .
In this precipitate , of which considerable quantities accumulate by the gradual solution of large qualntities of copper passing through the process , be found the following constituents : Antimony ... ... 9322 Iron ... ... ... ..o .
030 Oxygenl ... ... ... ... 2482 Arsenic ... ... ... 7-40 Nickel ... ... ... ... .
2260 Sulphr .
2-46 Platinum 04 ... ... 04 Cobalt ... ... ... ... 0 086 ) Selenium .
1-27 Gold ... ... ... . .
0 98 Vanadium ... ... 0-64 Sand .
190 Silver ... ... ... . .
4-54 Tin ... ... 33 50 Leacl ... ... ... . .
015 1 Copper ... ... 9-24 " Of these constituetnts , the ten first metals were obviously derived from the copper , in which they could have beenl scarcely detected unless by this accumulative process .
Of the remainder of the constituents , the tin in a great measure is derived from the solderings .
" The results obtained in the analysis of tne copper wires which you forwarded to me , appear to establish one fact in a satisfactory manner , viz. that the diminution of conductivity observed in certain specimens of copper is due to the presence in these specimens of a certain amount of foreign matter , and not , as it has been supposed , to a peculiar change in the physical condition of the metal ; for in the specimens analysed the conductive power rises in the same order as the total amount of impurities diminiishes .
" I have , &amp;c. , ( Signed ) " A. W. HIOFMANN .
" Professor Williamiz Thom8on , F.R.S. , 8c.9 ' It appears therefore that in the case of these four specimens , the electric conductivity is in order of purity of the copper ; but yet that only extremely small admixtures of other substances are to be found even in those which have but half the conductivity of the best .
P Petersburgli Acad. Btull .
vii .
p. 218 .
On the other hand , I have found by experimenting on artificial alloys , that comparatively large admixtures of lead , iron , silver , and zinc seem to produce sometimes improvement , sometimes little or no sensible influence , and sometimes ( as in the case of zinc ) an injurious effect on the conductivity of specimens of pure electrotype copper from which the alloys were made , The largeness of the proportion of other metal required to produce any considerable deterioration in comparison with that of the whole amount of impurities which Professor Hofmann 's investigation demonstrates in specimens of low quality as to conductivity , is worthy of remark , and seems to indicate that this low quality must be due to other than metallic impurities .
The great difference between the conducting qualities of two specimens of electrotype copper , from which two series of allovs were separately prepared , seems also to indicate some as yet undiscovered cause , as operative in general .
I am assured by Messrs. AIatthey and Johnson , by whom all the alloys were prepared , that similar m-lethods were followed and equal care bestowed to ensure purity in the two cases .
T'he resuilts of my measurenments of conductivity are showni in the following Tables : TABLE I.-Two Series , Nos. V10 and .
Nos. 1-32 , of Specimens prepared by Messrs. Matthey and Jolimison from pure electrotype copper , and the same alloyed with other metals , as specified .
S~~ft of compound .
8~~~pecific Specification o compouoduti _____________________ ~~ ~~~vity .
Pur _op e ... ... ... ... ... ... ... ... ... ... ... ... ... .
_ ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... __5 SERIES I : .
1 Pure copper ... .
138 52 Pure copper alloyed with *25 per cent. silvrer ... ... . .
138 53 Pure copper alloyed with *13 per cent. silver ... ... . .
139.5 4 Pure copper alloyed with *25 per cent. lead ... ... ... ... ... ... . .
144.5 Pure copper alloyed with:13 per cenit .
lead ... ... ... .146 ... ... ... .
j. 6 Pure copper alloyed with 25 per cent. tin ... ... ... ... ... 131 7 Pure copper alloyed with 13 per cent. tin ... ... ... ... ... ... ... 1..33 8 Pure copper alloyed with *80 per cent. zinc ... ... ... ... ... ... ... 1-25 9 Pure copper alloyed with 40 per cent. zinc ... ... ... ... ... ... ... 120'5 10 PLure copper alloyed with 1I40 per cent. ziInc. .
1 03 TA BILE L.-cotinnued .
Specific Specification of coilpoitnid .
coilductivity .
SE RIES II .
1 997 5 copper+2 5 silver ... ... ... ... ... ... ... ... .
69-8 2 9987 copper+1l3 silver ... ... ... ... ... ... 1-.-.-.- ... 117 73 997-5 copper+2-5 lead ... ... ... ... ... ... ... ... ... ... ... ... . .
945 4 998 7 co , , per-13 lead ... ... ... ... . .
105-8 95 copper ?
+2'5 in ... ... ... ... ... ... ... ... . .
, .
, .
, .
, .
, .
, ... .
91-6 0 998-7 copper I-3 tin. .
116-9 7 999 copper l silver ... ... ... ... . .
126(7 8 99 9 copper+ I lead ... ... ... ... ... .
1342 9 999 copper+-5 lead -5 sil-ver.128-0 10 Equal parts of I and 3 ... ... ... ... ... ... ... ... ... ... ... ... .
89 311 997 5 copper ?
-2 5 iron.129 7 12 958-7 copper+1-3 iro..113 7 13 1000 copper+2 5 protoxide of copper.122 5 14 1000 parts of 3 &amp; ?
+2 5 protoxide of copper ( too brittle to test ) 15 1000 parts of 4 &amp; +2 5 protoxide of copper.119-7 16 997.5 copper+2.5 ziiiC ... ... ... ... ... zinc .
, ... ... ... ... . .
108-9 17 995 copper+2-5 lead+25 ' ) zinc ... .
85.1 1c 9 9335 copper+2.5 lead , +2 5 iron .
131.5 19 998-7 parts of 11+1 3 lead.135 0 20 997 5 parts of 1:1+25 zinc ... .
77 6 21 998 7 parts of 11 +3 zinc ... ... ... ... ... ... ... ... ... ... ... ... ... 95-2 22 9i7 parts of II 1+3 lead &amp; -+ 13 zinc ... ... ... ... ... ... ... 117 6 23 992 copper+8 zine.118-9 24 99 , X6 copper+4 zinc .
117-0 25 986 copper+ 14 zinc.80.2 26 982 eopper+18 zinc.102 27 994 copper+6 zine.109-15 ) 28 980 copper+20 aluLrniniu.44-0 29 990 copper+10 alumiinium. .
I 128-7 30 995 copper+5 aluminiumi ... ... ... ... ... ... ... ... ... ... .
122-5 31 997 copper+3 aluminium. .
I. I ... ... ... ... ... ... ... ... ... .
130 2 32 PuLre copper , from wh-ichi all the above were made.120-9 TABLE IIL-First Series ( 10 specimenes ) arrariged in order of colnductivity .
|O*F ~~~~~~~~~~~~~~~Specific Specification of conmpounLd .
ciductif }t m -vity .
5 Pure copper alloy ed with *13 per cent. of lead ... 146 4 Pure copper alloyed with *25 per cent. of lead ... 1445 3 Pure copper alloyed with *13 per cent. of silver ... 1395 2 Pure copper alloyed with *25 ) per cent. of silver ... 138-5 1 Pnre copper ... ... ... ... ... ... ... ... ... ... ... ... ... [ 385 '7 PPure copper alloyed with -13 per cent. of tin ... 133 Puire copper alloy edl with -2t5 per cent. of tin 131 8 iPutre copper alloy ed with '80 per cent. of zine ... ... ... ... ... ... ... .
4 125 9 IPuire copper alloyed w , AI itht 40 per cent. of zinc ... 120t5 lO Pure copI n alnied with 140 per cent. of z.1nc0 ... ... ... ... ... 103 Pur cope aloedmiti TABLE II.-Second Series ( 32 specimens ) arranged in order of conductivity .
C. ) Specification of compound with manufacturers ' description conductiof mechanical quality of wire .
vity .
19 998 7 of No. 11 +13 lead : fair.135-0 8 999 copper+1 lead : fair.1342 18 995 copper+2 5 lead+2 5 iron : very good.1315 31 997 copper+3 alu:rminium : good.130 2 11 997-5 copper+2 5 iron : not very good ... ... ... ... ... ... ... ... ... 129 7 29 990 copper+10 aluminium : good ... ... ... ... ... ... ... ... ... ... ... 128 79 999 copper+-5 lead+-5 silver : rather better than No. 8 ... ... 128 ?
0 7 999 copperf-I silver : fair ... ... ... ... ... ... ... ... ... ... ... ... .- .
126-7 13 1000 copper+2-5 protoxide of copper : very bad ... ... ... ... ... 122-5 30 '995 copper+5 aluminiumn : very good ... ... ... ... ... ... ... ... ... ... 122 5 32 Pure copper : very good ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... 120-9 15 1000 of No. 4+2-5 protoxide of coppor : better than No. 14 , but not good ... ... ... ... ... ... ... . .
119-7 23 992 copper+8 zinc : first-rate alloy ... ... . .
118-9 2 998 7 copper+l 3 silver : fair , but rather frangible ... ... ... ... ... 117 7 22 997 5 of No. 11+1 3 lead+iJ3 zinc : very good indeed .
117 6 24 996 coppei ?
4 zinlc : moderately good ... ... ... ... ... ... ... ... ... ... 117 06 998 7 copper+ 13 tin : perhaps not quite as good as No. 5 .
1:16 9 12 998 7 copper +13 iron : frangible.1137 27 994 copper+6 zinc : good.1095 16 997-5 copper+25 zinue : first-rate alloy.108 94 998-7 coppei 1 -3 lead : rather better than No. 4. .
1058 26 982 eopper , 18 zine : very good ... ... ... ... ... ... ... ... . .
102 3 21 998 7o No. 11 +l3 zinc : very fair ... ... I ... ... ... ... ... 95 23 997 5 copper+-f2 lead : good , but requiires care ... ... ... ... ... .
94 55 997.-5 copper +2 ?
tin : mnuchl the same as Nos. 3 and 4 ... ... ... 91 6 10 Equal parts of Nos. 1 and 3 : bad , frangible ... ... ... ... ... ... ... 89-3 17 995 coppe +2 o5 lead+25 zinc : very good ... ... ... ... ... ... ... ... 85 1 25 986 copper+14 zinc : first-rate alloy ... ... ... ... ... ... ... ... ... ... 80 2 20 997-5 of No. 11 +2-05 zinc : very fair ... ... ... ... ... 77 61 997-5 copprr 2-5 silver : fair , but rather frangible.69-8 28 980 copper+20 aluminiumii : not very good ... ... ... ... ... ... ... 44-0 14 1000 parts of No. 3+251 protox .
copper ; alimost unldrawablel ( too brittle to test ) .
The alloys numbered 14 and 15 were prepared with a view to testing the possible effect of a suboxide of copper mixed or combined with the mass .
Although they do not seem worse than others of nearly the same metallic composition , it cannot be considered that they demonistrate that oxygen exercises no influence , as the portion of oxide itntroduced may have been reduced in the melting ; and indeed it is quite possible that some accident in the melting may possibly give rise to oxidation to a greater or less degree , and may cause some of the irreguilarities and uncertainties which have been observed .
( hi this I may remark , that althouigh I have fouiid that no mechanical alteration by hammnering , twisting , &amp;c. p)rodulces an'y consideiable effect of the conductivity of one piece of solid copper , I have not yet found whether or not specimens either good or bad retain their specific qualities after melting .
I may add , that it will be of gr eat importanice to ascertain the laws of variation of conductivity with temperature , of different specimens of nearly piire copper differing largely in conductivity .
I have hitherto used standards of copper wire in all the relative determinations of conductivity which I have made for different commercial specimens and artificial alloys of copper ; and before I found the very large differences of conductivity shown in this and in my preceding communication to the Royal Society ( June 15 , 1857 ) , it seemed natural to suppose that the relation between specimen and standard would remain constant , or nearly constant , when the temperatures of the two are varied to the same extent .
Now , however , it seems scarcely probable that this can be the case , and a rigorous experimental examination of the influence of temperature becomes necessary .
PIS .
April 11 , 1860.-I append the following extract from evidence which I gave on examination before the Government Committee on submarine telegraphs on the 17th December , 1859 , as it bears directly on the subject of the preceding article , and shows what degree of weight may in my opinioni be attached to the synthetical attempts which have been described .
( Chairman .
) Question 2458 .
Soon after you became a Director of the Atlantic Telegraph Company , was your attention directed to the conductivity of copper ?
-Yes .
2459 .
You instituted a series of experimenits , did not you , to determine the variation of this quality in different samples of copper ?
A number of samples of copper were , at my request , put into my hands for the purpose of measuring their conductivity in consequeilce of my having accidentally noticed differeiices greater than I expected in the conducting power of one or two samples which I had had previously .
2460 .
Will you be good enough to state the genieral results at which you ultimately arrived , arid your modes of experimenting , ?
My modes of experimenting did not differ materially from the methods which had been followed by certain other experimenters , especially in Germany , and were in reality all based on Professor Wheatstone 's irnvention of a beautiful method for comparing resistances , to which I have frequently referred as Professor Wheatstone 's electric balance .
2461 .
What were the results at which you arrived ?
-Tbat different specimens chosen at random from the stock supplied for manufacture differed immensely in conducting power .
2462 .
Although nominally the same quality of copper ?
-Yes , although nomilnally the same quality of copper .
All those specimnens of wire were supposed to be of the very best quality , the only copper supposed to be good being that which admitted of being drawn into wire suitably for the purpose .
A good mechanical quality was necessary to prevent frequent fractures in the wire-drawing ; and to undlerstand that , I should say that hanlks in unbroken lengths amounting to a large mass were always required , the worse metal being found to break before it could be drawn into a hank of a certain size .
The mechanical qualities seem to have been satisfactory , but Ino suspicion whatever was entertained that there were also large , e differences in electric conducting power .
W. Weber had many years before pointed out considerable differences in different specimens of copper wire which he had tested .
I found diffeerences much exceeding those , and I did not , as I expected , find any approximation to a uiniform average amoing the di-fferent specimens tested ; some specimens I found nearly double in their condutcting power , compared with others , reckoned according to the weight and length , allowing for the variations of gauige .
Calling the best specimen which I had in the summer of 18357 , 100 , I found many specimens standing at 60 in specific conductivity , many standing at 50 , many standing at 80 , a few above 90 ; and so far as I can recollect , the average of a large number of specimens that I then examined may have stood between 60 and 70 , but I consider the statemenit of such an average to be of no value , it is so much a matter of chance .
If I had received a dozeni specimens of a low quality below the average , or if I had chanced to receive a dozen specimens of a higher quality , the average would have been so much the low er or the higher .
I never hta(d anl opportunity of measuring the conductivity of 200 or 300 miles of submarine cable ; such alonie would have given me exact information as to the average for that por'tionl of cable .
I may mention that a inonth or two later , still in the surmlmer of 1857 , 1 received specimenis of wire which were in stock for submarine telegraphs , -for some of the Mediterranean telegraphs , I believe , ---which stood as low as 43 on that scale ; and , lastly , I may menition that I have since met with specimens standing 2 or 3 per ceiit .
above the 100 ; and an artificial alloy , which I had prepared , stood , so far as I can estimnate , as high as 11 1 .
2463 .
What was that alloy ?
-The alloy consisted , so far as I can recollect , of copper and 13 per cent. of lead .
I have made experimenits upon a series of alloy1s , in all about 43 or 44 , and have recenitly repeated the examiinationso as to arrive at accuracy , within certain limits ; and I expect , immediately , to be able to communiicate to the Royal Society , for publication , the results .
A few months ag , o I sent a provisioinal list of the specimens , showinthe relative coniductivity of those alloys , bhut , possibly , requiring co'rrection1 as to the absolute coniductivity stated .
That list was conmmunicated to Mr. Latimer Clarke , and , I believe , a copy of it was laid before the Comynittee .
2464 .
( Professor ftheatstone .
) Wrere you qtuite certain that vou employed pure copper in your experiments ?
-I couldl not be quiite certain .
24651( ) .
The copper might be alloyed with other things than metals ; is it not very probable that it might contain some suiboxide , and that the mixing of lead afterwards with it might have reduced the suboxide , and therefore have given it a higher coniducting power on that accouint ?
-That is possible .
I cannlot say that I am at all satisfied that the experiment.s which I have made point out distinctly the relation between the ascertained chemical combiniation and conductivity .
I may mention that one of my alloys was made with a suboxide melted with the copper ; hut the uncertalinty of the process of melting the suboxide and the uncertainty as to how much of the oxidation may have disappeared in the melting , prevented me from attributing much weight to the experiment .
2466 .
( Chairman .
) What was the result with that alloy ; was it a low result , or a high result ?
-A moderate result ; not a low result .
2467 .
But not a high one ?
-A somewlhat high result ; but I may mention that in one series the higlest condctCtiVity was foutind with a mixture of lead qind iron ; fractionis of a per cent. of lead , and fractiouls of a per cent. of ironi miixed with pure copper gave a hligher conductivity than a lnomlinally pure copper , with which the alloys were prepared .
I must mention further , that in two series the alloys , both prepared by Messrs. Matthey and Johlnson , and as I have been assured with equal care , gave results preselnting conisiderable discrepancies ; the conductivity of the pure copper in the first stood high , nearly agreeing with the 100 of mriy first scale , the pure copper of the seconcl series fell considerably below that limit .
On this account it appears that even pure copper , carefully prepared by the electrotyp)e process , does not always give us results which show perfectly in point of conductivity ; but to make such experiments in a satisfactory manner , it would be necessary to have a thorough chemical investigation , both synthetical and analytical , of the metals used ; such a thorough investigation I have not been able to carry out , in consequence of the large expense which it would entail .
I may mention that AIr .
Matthiessen has gone through a series of experiments oln alloys , of which the chemical composition has been ascertained with all possible accuracy , and has , I believe , arrived at highly important results relative to electrical conductivity .
I have been in coimulnication with him , and have supplied hiim with a specimnen of one of my standards .
Ile mentions to me that he has obtainied specimens conducting better to a considerable extent than the 100 of my first scale .
In that respect he has confirrned what I have myself ascertained , having myself found specimnens as high as 111 on that scale .
A number of alloys of definite chemical compositioni , prepared with great care by Mr. Calvert of Manchester , and already tested by him for therrmal conductivity and for mechanical properties , have been put inlto myiy hands , in order that I may measure their electric condmmctivities .
I hope soon to be able to obtain and publish results for this series of alloy s.

