CLVI. 011 Electricul Eudosniose. By Mr. JAMESNAPIER. HAT two dissimilar solutions separated by a porous par- T tition will pass the one into the other is a pliaerioriienon long observed the only necessary condition being that the liqiiids have a strong tendency to cotnhine and that the one is more capable of entering into or wetting the porous par-tition than the other. The liquids eminently fitted to effect this are alcohol mid water and saturated solutions of sonie salts and pure water. But the phenomenon of endosniose takes place also when an electric current passes through liquids separated by a porous partition. When all the above con- ditions are absent aid even wklen these conditions are pre- sent the endosmotic current will follow the electric althougfi in ;1 cotitrary direction to that which would take place were Mr .J.Nap ier 011 E'lrctrictrl EtJ closmose. 29 there no electric current passing showing that under these circumstances it has its origin in the passage of the electric force. This fhct was first matle known by Mr. Porrett in the Annals of Philosophy for 1816. The object of' iliis paper is not to define the cause but to point out the different conditions of electrical endosmose and the important part it plays in electro-chemical investigation. The conditions under which electrical endosinose are ob-served are that the two metals constituting a battery or the two electrocles of a battery be placed in separate vessels or divisions of the same vessel ; one of the vessels or partition being composed of a nxtterial suficiently close in texture to prevent the mixture of'the two liquids and porous enough to allow the electricity to permeate through it.The substances generally used are bladder parchment unglazed porcelain &c. ?'tie last is what I have generally used in the experi- merits to be detailed. ,. 1he general effects of electrical endosmose are a portion of the psitive solution passes along with the electric current into the negative solution not by electrolytic action as has been generally supposed but by endosmotic action ; and this endosiiiotic current is coiifined to the direct influence of the electric current or facing the metals composing the electrodes or battery. There art some circumstances in which the posi- tive solution gains in quantity making an apparent curreiit contrary to tlie electrical; these will be referred to as we proceed.Electrical endosniose manifests itself in two ways or rather is of' two kinds which may be distinguished as the mensurable and the iwimcasui.czbIe the fornier being the result of the trans- fer of' water fimni one cell to another the latter of'a salt or acid lielcl in solutioii and which is being decornposed; this rnay take place to a11 extent equal to the ~iiole salt lieltl in solution in the positive cell without the quantity of' the two solutions being niatei.ially altered ; mid that this transfer is front eiidosinose will be best illustrated by de~iling a few exper i in e11 t s stati r 1g how e ve r that the invest igat ion n ecess ar i 1y involves :t pat nuiiiber of experiments the whole of which I cannot detail and the conclusions come to have been the result ofthese.It may be stated here that the vessel or divi-sion in which the positive electrode or metal is placed is termed the positive solution and the negative division the negative solution. 1. 20 ounces of water in which were dissolved 600 grains of sulphate of copper were inade the positive solution arid other 20 oiinces of' water with 500 grains ofsulpliate of cop- Mr. J. Napier Electrical Errdosmose. per were made the negative soIution. A copper electrode previously weighed was put into each of these solutions and connected with a Wollaston's battery of nine pairs the current allowed to pass until the negative solution was completely exhausted of copper which required sixty-four hours the battery not renewed during the experiment.At the end of' the experiment the electrodes being again weighed the posi- tive had lost 257 grains the negative had increased 250 grains. The positive solution had lost 6 ounces by nieasure the iiega-tive had gained 4 ounces. The greatest portion of this change took place within the last ten hours. The 2 ounces lost are iiiostly from evaporation. The positive solution was evapo-rated and crystallized and there were obtained 905 grains of sulphate of copper 95 less than was originally dissolved ; but the partition being saturated with salt may account for a por-tion of the loss.The 500 grains which were originally in the negative solution would only contain 127 grains of metallic copper; but there is nearly double of this deposited showing that the salt had been transferred by some means fiom the positive to the negative in quantity amounting to nearly the whole of the salt originally dissolved in the positive solution. 11. 100 grains of sulphate of copper were dissolved in 4 ounces of water and put into each division in the same man-ner as the last experiment. The partition in this case was much closer in texture than the other; the same battery power was used and continued till the negative solution was exhausted which was nineteen hours. .The negative electrode being weighed it had increased 34-4grains; the negative solution had increased one-fourth of a11 ounce; the positive solutioii had lost three-eighths of an ounce.Comparing this experiment with the last there is a great discrepancy the only known cause being tlie closer texture of the diaphragm ; the transfer of the salt is only in the proportion of one-third that of the former and the measurable endosniose is still fur- ther out of proportion being only one-sixteenth ; while the amount of decomposition is only one-seventh of the 500 grains. The electrodes in both experiments were the same in size. The question now suggested itself whether tlie transfer of the copper salt from the positive to the negative cell was the result of endosmose or of electrolysis corresponding with the results of the late Professor Daniel1 and Professor Miller who supposed that certain bases underwent electrolytic transfer in fractional proportions and that these proportions might vary according to the texture of the diaphragm or that kind of' peculiar resistance given to the passage of a current when passing from one solution to another.Mr. J. Napier on Electrical Ihidosmose. In order to determine which was the true cause the nega-tive cell was charged with a weak solution of caustic potash the positive being sulphate of copper ; gas was freely evolved from the negative electrode; in a little time the porous dia- phragm facing the electrodes became coated with oxide of' copper which greatly retarded the current and niade it neces- sary to use ti more powerful battery.After twelve hours the oxide which had accumulated upon the diaphragm and at the bottoni of the positive cell was carefully collected washed dried and weighed 7'3 grains. The potash of the positive solution being neutralizecl by muriatic acid had a solution of chloride of barium added which gave a precipitate of sulphate of bwytes weighing 26 grains equal to about 9 grains of sul-phuric acid. From this it appeared that the copper had passed from the positive to the negative as sulphate of copper and therefore not by electrolysis. In repeating similar experiments with very dilute alkaline and earthy salts in the negative cell I have observed that the oxide of copper fornied being a conductor of electricity it often floated between the diaphragm and negative electrode as a kind of spongy fibre and condiicted the electricity through it as a solid constituting its extremity in connexion with the porous partition into the electrode the result being the reduc- tion of the oxide in and upon the diaphragm closing it up with metallic copper similar to that observed in the operations of electro-metallurgy when any of the electrodes are allowed to touch the diaphragm.It Iias often surprised me how nii-nute a fibre would connect the electrode and cell and produce a wide-spreading result. A solution of cyanide of potassium was next tried in the negative cell having sulphate of copper in the positive; the current passed four hours. The cyanide solution smelt strongly of hydrocyanic acid ; the solution was found to contain both copper and sulphate of potash ; the quantities were not de-termined.The next object was to use two salts of such a character that if that fi-om the positive cell passed to the negative by en- dosinose it would remain in solution ;if' by electrolysis the base would be reduced to the state of an insoluble oxide. The negative was accordingly charged with a solution of caustic potash and the positive with cyanide of copper and potassium the latter in excess After eight hours during which gas was freely evolved from the negative electrode and upon it was a sniall portion of copper powder no oxide was formed ; the solution contained cyanide of potassium and copper the latter being precipitated b~ hydrochloric acid and fused gave Mr.J. Napier on Electrical E)ido.srnose. 3'5 metallic copper with strong smell of hydrocyaaic acid. From this it appeared that both the copper and potassium salt were transferred by endosmose. The double cyanides of po-tassium and silver and of potassium and gold were next sub- stituted for the copper in the positive cell with similar results. Nitric acid was also substituted for the caustic potash iii the iiegative cell with the double cyanides of copper silver and gold iii the positive which would give the contrary result of last experiments namely -precipitating the metals if tram-ferred by endosmose Init dissolving or holding them in soh -tion if transferred hy electrolysis. In a short time in eacii experiment the surtace of the porous diaphragm facing the negytive electrode became coated with the cyanide of the metal which accumulated and dropped to the bottom of the vessel.At the termination of the experiments the acid solution smelt strongly of hydrocyanic acid and contained much nitrate of potash. With the gold aird silver salts the acid solution had not the slightest indication of their presence in solution but with the copper there was a corisiderable portion present which was owing to the cyaiiide of that metd being clecom-posed by nitric acid ; but this experiment being repeated with the copper cyanide in the positive and sulphuric acid in the iiegative no trace of copper in solution was found in the acid. ,7 1 he next experimmt was to determine if any portion of' the potassium of the potash salt was transferred by electrolysis ; tor this purpose two porous vessels were employed one filled with n solution of cyanide of' potassium the other with dilute nitric acid; these were placed in ii crlass Yessel containing a ? solution of nitrate of silver ; the positive electrode was placed in the cyanide of potassiurn the negative in the nitric acid.The cyanide of potassium which passed from the positive cell to the silver solution in the glass vessel was dec*oniposed pro-ducing cyanide of silver and nitrate of potash. After twelve hours the cyanide of silver formed was carefully collected washed and dried ; it weighed 65.6 grains ; the rernaiiiing silver solution had as much muriatic acid added as precipi-tated the silver.The clear filtered solution was now evapo-rated to dryness and kept for some time at the point of fusioii. There were obtained 28 grains of nitrate of potash which is 4 grains less than the equivalent of' cyanide of silver obtained ; but the nitric acid in the negative cell also contained a little nitrate of potash and silver which had passed from the glass vessel and will account for this loss. From these and a vwiety of other experiments of a similar kind with different salts I consider that no base of an electrolyte is transferred by electrolytic action but that salts being electrolysed are dl MI.. 33 J. N:rpier on Electrical Endosrrwse. transferid more or less from the positive to the negative elec- trodes by entlosmose aiiti that the amount of this varies nc- cording to the texture of the porous diaphragm the power of the electric current and \Jai*ious other niodifjing circuin- staiices which will be apparent as we proceed with the inquiry irito the cause and circutristances of electrical endosmose.111 refkrring to the first experiment it was observed that the greatest amount of nieasurahie endosniose took place during the last ten tiours when both solution and battery were nearly exhausted. 5OOgrains ot‘sulphate ofcopper were again put into each cell the solutions accurately nieasured and the current of9 pairs passed for twenty hours. ‘l’he deposit on the negative electrode was 116.3 grains ; this solution had increased in measure three-eighths of an ounce the positive had lost half an ounce ;being evaporated and crystallized the negative gave 226 graiiis crystals the positive gave 749 grains.In this ex- perinleiit we have 184grains of the salt transferred and only three-eighths of ail ounce of iiieasurable eiidostnose. 960 grains of sulphate of copper were dissolved in 20 ounces of water and put iiito the negative cell while the positive was charged with dilute muriatic acid. A battei-y of 18 pairs was connected and kept in action twenty-four hours; it was then found that the negative electrode had increased in weight 178 grains tiiicl tlie solution had gained in measure Z+ onnces. A small portion had nitrate of silver added which gave a slight niilkiiiess but not amounting to a precipitate.The whole so!u-tion of the negative cell being evaporated and crystallized gave sulpliate of copper 463 grains. This shows a transtLr ot’sul-pliate of copper from the positive solution eqml to 162 grains. r? Ihe positive solution was of a deep blue colour it had lost I)y measure 24 ounces; the electrode was covered with a white powder and had lost 212 grains. Here again the two kiiicls oferitiosmose are perfectly distinct. But we have another curious result viz. the muriatic acid apparently refusing to be tr:iiisiki*red from tlie positive to the negative solution. ‘l’his attracted particular attention and the next experiment was coiicluctecl in the following manner :-lo0 grains of sulphate of copper were dissolved and put into a small porous vessel the solution measuring 2; ounces ; this was supported at the top ofa deep glass vessel filled with dilute muriatic acid so that several inches of solution were under the porous vessel the surface of the two solutions beiiig level ; by t(iis arrangement the copper salt formed at the positive electrode by the cur-rent sunk to the hottow by its own gravity and remained un-der the porous vessel so that no transfer of salt could take Chenz.Suc. Mem. VOL.III. D Mr. J. Napier on E:lcccti*icalEudosmost. place. Two copper electrodes were used and the curreiit continued until the negative solution was exhausted of copper. When this was completed the negative electrode had gained in weight 25.1 grains the solution had increased in measure half an ounce.On testing this by nitrate of silver only a slight milkiness was obtained. The positive solution was deep blue at bottom but only a slight tinge above the bottom of the porous vessel ; the positive electrode was coated nearly to one-sixteenth of an inch with a white pasty matter a great portion had also fallen to the bottom of the vessel the elec- trode had lost in weight 132 grains. Here we have rnea-surable endosniose amounting to half an ounce without any transfer of the acid in solution. I now took two porous vessels the one charged with ounces by measure of a solutioii of sulphate of copper the other with dilute muriatic acid ; these were placed in a glass vessel filled with a measured quantity of distilled water the cells placed 1 inch apart ; two copper electrodes were used the muriatic acid cell made the positive solution the sulphate of'copper the negative.A current from 9 pairs was kept up for eighteen hours. At first the decomposition was exceedingIy slow but it afterwards increased ; the results of this experi- ment were-Positive solution lost three-eighths of an ounce pole covered with white powder and had lost in weight 37 grains; solu-tion contained sulphate of copper. The solution in the glass vessel was tinged blue by sulphate of copper had lost in mea- sure l+ ounce and gave a copious precipitate with nitrate of silver showing a transfer of muriatic acid from the positive cell. The negative cell had increased in measctre'l ;ounce.Nitrate of silver gave no indication of muriatic acid the electrode had increased in weight 26 grains. A similar experiment was again repeated the negative cell and glass vessel being both charged with distilled water the positive with dilute muriatic acid ; platinum electrodes were used. A 9-pair battery was attached for sixteen hours. The current passing was sufficient to keep deflected a galvano-nieter needle but no evolution of gas was observed till nearly the end of the experiment when the poles became covered with sriiall bubbles of gas not large enough to be evolved. The results of this experiment were positive cell lost in [Ilea-sure one-eighth of an ounce. The solution in the glass vessel was slightly acid giving a precipitate with nitrate of silver and had lost in measure 1; ounce.The negative cell had increased 2; Mr. J. Napier OTL Electrical Endosmose. in measure 1 ounce did not change the colour of blue litmus paper nor give any precipitate with nitrate of silver here also as in the last exprrimeiit we have muriatic acid being transferred in small quantities from the positive to the inter- vening liquid. But another feature preserits itself the great amount of measurable endosirlose and the almost entire absence of electrical decor~iposition ; at the sanie time the en- dosmose being principally confined to the two vessels contain- ing water as if the acid in the cell had only acted the part of a11 electrode. On reversing the condition of the last expe- riment the positive and glass vessel being charged with di- stilled water the neptive with the dilute muriatic acid 9-pair battery for sixteen hours the positive vessel lost 1; ounce; the glass vessel had increased ti ounce while the negative or acid solution underwent no alteration showing again the en- dosniose only between the two cells containing water.Several experiments were repeated both with the single and double cells having muriatic acid as the positive solution with sulphate of copper water and alkalies as the negative so-lution; in some I had transfer of acid in small quantity in others no indication of transfer; I believe the cause of dif-ference to be in the texture of the partition ;wheii R diaphragm was used less porous the enclosniose was considerable in one case amounting to 4 grains in thirty hours.Nitric and sul- phuric acids which are much more easily transferred than muriatic acid have also had their transfer resisted by two closely-textured diaphragms passing through one into the mid-division but not into the negative solutioii :however these results show that all kinds of salts or acids are not transferred with the same facility probably from their power of contluc-tion. While operating with the double cells having an acid in the positive and water in the middle and negntive cells I have several times observed that both positive and negative solu-tions increased in bulk at the expense of the middle solution ; in one iiistance the positive solution had gained I ounce the ne- gative gained three-fourths of an ounce and the middle solution lid 1ostl;ouiice.The negative and positivesolutionsoriginally contained only 2 ounces the electrodes measured 2 by i inches; ~ievertheless in these instances the transfer of the acid from the positive to the negative was considerable. The different acids gave different results of this sort; with sulphuric acid the increase was greatest muriatic acid one-half less nitric acid about one-fourth of the sulphuric ; whether this be the result of ally fixed law I cannot say the times I have ob-Dg M r. J. Nnpier 011 Ekctricrd Eridosmose. served it king too few. As I have never observed these re- sults except with the acids I am inclined to think the cause to be what lias been already stated namely the solution act- ing as the electrode thus exposing a surface of acid to water iii a state of excitement or tension suacient to attract by its exalted affinity that fluid and cause an increase of bulk in the cell in a similar manner as I shall endeavour to show in a forthcoming paper to that which caiises a greater amount of action at the positive electrode than at the negative in any electrolyte and with muriatic acid and copper as already re- ferred to in a former part of this paper amounting occasion- allvI to five times that of the negative.shall now give the results in a tabular form of a few ex- periments made to determine the relation if any between the m eas11rable en dosin ose and eI ect ri c a1 decomposition .The experiments were all made with the double cells of three compartments ; the battery used being Wollaston’s of 9-pair intensity ; the time of each experiment was sixteen hours; the electrodes of platinum. The acids used were di- luted as 1 to 25 water; solution of sulphate of copper the same strength in all. .mount of Positive Amount of deposit measura-solution. diate). 1 on decomposition. de endos-~~ ---I Ounces. I-’ Ounces Grains. SO~ Water loss + I 4 lot measurable mose. Water Water so3 + none 8 ......... ~05 Water Water L loss 3 5 ......... Water Water KO5 la gain Q none ......... HC1 Water Water 8 loss 1% 1 ......... M’ater Water I HC1 I+ gain l$ none ......... Water Water gain 8 5 1 Water Water gain + 4 10 so3cuo SO3 gain + 2 33 so3 so3 gain 6 1 P 24 SO3CuO Water loss + I 7 11 s0” Water loss 1 ~ 14 30 IICl Water loss 1; 12 27 so”cuo l03c11c loss 53 IICl IICl so3cu & gain f A-43 SO3 k.1’ate r Water lossgain 3+ loss 9 16 .?ot measurable Water Water so3 loss 1 1% evolution of gas IICl Water Water gain+ loss 1$ 1 not measurable K 0s Water Water gain & loss + + not measnrable so3 Water Water loss ij loss 3 8 riot measurable Water Water none loss 1 1 not measurable II The following table is of the same kind as the last but the time of action and the power of the battery varying as stated.Mr. J. Napier 011 Electrical Erdosmose. -Negative Loss ir Change in Gain solution. positivl glass vessel in ne Power of battery.gative Grs. lour SO3CuO none loss i'Tsoz r'i oz 21 &oz 2 2 I). large* pls. ~03cUo+z loss AOZ + 02. 12 + 02. 2 IIlttO jO3CuO none none none 23 none 2 Ditto ;O"uo ... ... ... 9 ... 2 4p. small pls. ;o CUO ... ... ... 15 ... 2 Ditto ;03cu0 ... ... ... 16 ... 2 Ditto io3cuo ... ... ... 8 ... 2 1 p. large pls. i03cu0 ... ... ... 7 ... 2 Ditto ;o"Cuo ... ... ... 8 ... 2 Ditto 103cu0 ... ... ... 6 ... 2 1p. small pls. O~CUQ' .,. ... ... 5 ... 2 Ditto 103cu0 ... ... ... 7 ... 2 Ditto ,o3cu ... loss +oz. + 02. 10 8 02. 4 Ditto O"u0 $02. gain + oz. + 02. 27 4oz. 4 9 p. small pls. * In those marked large plates the zinc plate of battery measured 6 by 4 inches ; in those marked sniall ziiic measured 2 by 2+ inches. The result of these esperiments as well as those given be- fore in detail shows that there is 110 relation between the rnea- stirable entlosmose arid the aniount of clecomposition in the cells; so that the two phaenomena must depeiid upon sonie-what different causes which now became an ohject of inquiry.Professor Faraday in his researches mentions that a current of electricity may be made to pass through soliltions without deconiposing them and Mr. Sturgeon gives it as his opinion that there is always an undecomposing current passing with a deconiposing current. These observations being applied to the results of the present inquiry into nieasurable and unmeasu-rable endosinose seemed to throw some light upon the distinct character of these pha3nomena for as will be observed the measurable endosniose seems to be greatest when the current has the greatest difficulty to pass through and when the de-coniposition is least ; and on the contrary the unmeasurable entlosmose is greatest when the battery is powerfiil and the curreilt passing fieely or rather decoinposition going on freely.In order to compare the two I now tried a few experiments with different powers ot' battery under constant circumstances as regards the clecomposition cells. Ihch division of a decom-position cell hat1 put into it 100grains of sulphate of copper dissolved in 3 ourices of water; copper electrodes were used atitl the current allowed to pass until the negative solution was exhausted of copper. The following is the mean of several trials.FI'ith a I-pair battery the negative sulution exhausted in 38 Mr. J. Nnpier OIZ Elect?-icciI E?zdosmose. forty-one hours deposited upon pole 29 grains being 3.4 grains more than the equivalent of 100 grains of sul hate which is equal to 8'5 grains of sulphate which have pas Ped by endosmose not including water of crystallization as I believe the salt passes without water. The negative solution had in- creased in bulk 1 ounce. With a %pair battery the negative solution exhausted in nineteen hours deposited upon pole 32 grains being 6*4 grains more than the equivalent of 100 grains of sulphate and equal to 16 grains of sulphate passed by eiidosmose. The solution had increased in bulk five-eighths of an ounce. With a 4-pair battery the neptive solution exhausted in twelve hours deposited upon pole 35.6 grains being 11 grains more than the equivalent of 100 grains of' sulphate and equal to 27'5 grains of sulphate passed by endosmose The solu-tion had increased half an ounce.With a 6-pair battery the negative cell exhausted in seven hours deposited upon pole 39 grains being 14-4grains more than the equivalent of 100 grains of sulphate equal to 96 grains of sulphate passed by endosmose. The solution in- creased in bulk three-eighths of an ounce. The two divisions of the decomposing cell were now charged with distilled water; two platinum electrodes were used ; the current was also made to pass through a solutioii of sulphate of copper to ascertain if suficiently strong to eflect any de- composition ; but in none of the experiments was any deposit obtained.Each experiment was continued thirty lioiirs when r the cells stood thus:- Witti one pair positive lost 1 oz. negative gained $ oz. ... two pairs ...... I oz. ... ... $ 02. ... four pairs ...... 1 oz. ... ... 6 02. ... six pairs ...... oz. ... ... 1; oz. ... nine pairs ...... 18 oz. ... ... 16 02. With these and all other experinlents with wiiter a similar vessel to the decomposing cells was placed alongside filled with water to note the loss by evaporation which in this case was three-eighths of an ounce accounting for the loss in the two cells above. A similar experiment was made with thirteen pairs of a Grove's battery the current passed four hours ; a gentle flow of gas was evolved from the electrode.No copper solution was used in this experimeut. The cells stood thus positive lost 2 ounces negative gained 12 ounce. The two divisions of the decomposition cell being again filled with distilled water into one was put a piece of zinc and into the other a piece of platinum connected by a wire; Aiiaiyis of n CoGctLt Orecfbund in IC‘estcrn Indict. 31) in tijrty-eight hours the zinc cell had lost three-quarters ofan ounce the platinum cell had gained half an ounce. This experiment was repeated many times with similar results. Two large cells were filled wiili distilled water and a piece of zinc measuring 4 hy 6 inches carefully weighed was put iiito one division and a piece of copper of the same size was put into the other division the current passed through a deli- cate galvanometer which kept deflected about 3’; the posi- tive solution was kept at a given height the negative was taken out as it increased keeping the two solutions as nearly level as possible; this was kept up for forty days when there was found to have passed through from the positive to the negative a bulk equal to 32 ounces allowing for the evapo- ration which was known by a similar vessel placed alongside.?‘he zinc was covered with a gray filrn and had increased in weig!it 12 grains; this being carefully dissolved off by am-monia water and the zinc again weighed it was found to have lost 36 grains which we inay take as the amount of oxidation during the experiment.The general conclusions which may be drawn from these experiments respecting endosmose are ( 1.) ‘l’hat the current of positive electricity passing through a liquid is always accompanied with a current oftlie liquid in the same direction. (2.) If the liquid contains ft salt or an acid that is under-going decomposition the ericlosinotic current is principally if not wholly confined to that salt or acid unaccompanied with water and therefore adds little or nottii~ig to the bulk of the liquid into which it passes. (:L) When the quantity of electricity which a battery is ca-pable of giving off is greater than the salt or acid can conduct the extra quantity if we may so term it passes through the water taking with it and thus inducing a flow of that liquid iuto the negative cell increasing the quantity ;the same effect being produced with water when no salt or acid is in solution hence the well-known fact that endosmose is greatest with pure water and even with currents that give no apparent de- coniposition or rather a decomposition so minute as to be un-observed.