60 MUIR AND SLATEH INFLUENCE OF WATER TI.-CONTRIBUTIONS FROM THE LABORATORY OF GONVILLE AND CAIIJS COLLEGE CAMBRIDGE. S o . 1K-00’1~ the Iuflueiice exerted upon the Course of Cejtcx,in Chemical Changes by Variations i n the A m o u n t of IVutey of Bilutiou. By 1\f. %I. PATTISON J I U I R Caius P r d e c t o r i n Cliemistry and CHAS. SLATER B.A. Scholar of St. John’s College. 1. In a paper by one of us (this Journal 1879 Trans. p. 311) an experimental examination was made of some of tlie conditions which affcct the equilibrium of the system CaCI + Na2C0 + zHLO. I t was there suggestted t h a t a study of tlie changes undergone by this sgsteni under conditions such t h a t the action of the various forces might be more disentangled from one another would he advisable.The present paper cmtains the results of experiments undertaken with the object of determining the influence on the equilibrinrn of the system of variations in the quantity of water present. 2. The solutions use2 contained respectively 320.7 mgrms. CxC1 in 10 c.c. and 305.54 mgrms. NazCO in 8.1 C.C. The sodium carbonat PLATEL (TABLE 1) CaCI and Na&O,. 1 1 molecules. 30minutes. l.8” Brrison & Sons. Lith. S Martins Lane,JK C OX T€IE COURSE OF CERTAI?; CIIEJIIC-IL CH-1NGES. 61 was added to the calcium chloride solution and the whole was then shaken up. Initial condition was CaCl Na,C03 H,O = 1 1 330 mols. Each 25 C.C. water added = 470 mols. Time = 30 minutes. Temp. = 16-18'. TABLE I. Mean percentage Water added. of CsCO3 In C.C. I n inols.produced. I 0 330 97.8 25 800 96.3 50 1270 96.0 95.1 75 1740 95.5 95.7 100 2210 93.8 94.6 200 4090 91-2 91.2 300 5970 $6.3 85.9 400 7850 $5.8 85.8 500 9730 82.1 i8.7 Actnnl numbem obtained. 97.5 96.8 96.03 95.4 93.1 91.2 85.9 85.8 78.7 98.2 96.1 96.03 96.03 96.03 96.08 9.3% 95.8 94.3 93.3 91.9 87.1 85.6 85.3 Another solution was used after this point. 10 C.C. CaC12 = Initial state = 1 1 2700 mols. Each 25 C.C. water added = 55.5 mgrms. 2700 mols. Water aacled. Mcnn percentage of &tual numbers I n C . C . I n iiiols. C'uCO produced. obtained. 100 12.500 75.0 175 20,600 66.5 66.0 67:O 200 23,300 62.2 62.2 225 26,000 52.0 52.0 The results contained in this table are graphically represented in the curve of Plate I.3. Thme results show that in the special case considered the amount of chemical change decreases regularly as the amount of water of dilution increases. The water which is added a'ppears to exert a purely mechanical influence on the change under consideration. We may suppose that the chances of collision between the mole-cules of sodium carbonate and calcium chloride are decreased the greater the number of inactive water molecules interposed. 4. I n the second series of experiments which we conducted the influence of water on the reaction SrC1 + H2S04 = SrS04 + 2HC1 was studied. The results obtained meTe very discrepant and seemed to allow of no reasonable geiiernlisation being deduced. These results will be referred to in the seqnel. 5. The reaction BaCl + R2C2Q4 = RaC,O + 2KCl xppcared t 6 2 BlUIR XSD SLXTER INFLUESCE OF W-iTER present a chemical change well adapted for the study of the influence of dilution.Barium oxalate is tolerably insoluble in water and the amount of change can be quickly and accurately determined by nieasuring the Quantity of undecomposed potnssiutn oxalate by titration with stan-dardised permanganate. The experiments were conducted by adding a measnred volume of potassium oxalate solution to the proper quantity d € solution of barium chloride previously diluted with a measured volume of water shaking up and then allowing to remain a t rest. At tlie close of the time allowed for the change to proceed a portion of the clear liquid was drawn off and run through a dr7 filter into n dry heaker and the amount of potamium oxalate in solution was determiried in an aliquot portion of this filtrate.TABLE 11. 10 C.C. BxCI solution used = 627.04 mgrms. BaC1 9.35 C.C. K,C204 Initial condition = BaC1 K,C,O H,O = 1 1 360 mols. solution used = 501.03 mgrms. Each 25 C.C. water added = 460 mols. 16-18'. Mean percentage of Water added. BaC1 decom-I n C.C. In mols. posed. 0 360 90 2 25 820 88.5 5 0 1,250 8 7.6 100 2,200 79.8 250 4,960 70.3 300 5,880 69.1 350 6,800 63.2 400 7,720 56.8 600 11,400 46.0 200 4,040 74.8 Time = 30 rnins. Temp. = Kumbers actually obtained. -89-09 91.32 88.2 88 s 57-5 87.8 79.1 80.2 79.3 79.9 75.0 74.5 70.4 SO.1 70.0 67.8 69.5 62.8 63.5 57.5 57.5 55.4 43.0 49.0 These results are represented graphically in Curve ,4 Plate 11.6. The generalisation of par. 3 may me think be also deduced from these results. The amount of chemical change decreases regularly as the amount of water of dilution increases. 7. Another serie3 of experiments was conducted similar to the fore-going save that the beakers in which the change proceeded were surrounded by ice TABLE 111. Details as before. Temp. = 3". Time = 30 miiis. Water zdcled. I l c m pcrccntnge of In C.C. In iiiols. UaCl; decomposrd. Numbers actually obtained. 0 360 25 820 100 2,200 200 4,049 3.00 5,860 400 T,720 coo 11,400 7wo 13,240 These results are represented grapkically in Curve B Plate T I . 8. In the nest series of expwiments we made the time longer still keeping the temperature low.TABLE ITT. Details 3s hefore. Temp. = 3". Time = 98 mins. Water added. In C.C. In moh. 2 3 F10 100 200 300 400 500 GOO i00 800 820 1,280 2,200 4,043 5,880 7,7 20 9,560 11,101) 13,240 15,080 Mean percentage of BsCI, 98-1 92.0 83-9 71.5 63.7 44.6 30.7 20.1 24.4 decomposed. 9". r / 3 These results are represented graphically in Curve C Plate IT. 9. We now raised the temperature keeping the time the same as before 6 4 MUIR AND SLATER INFLUENCE OF WATER TABLE V. Details as before. Temp. = 18". Time = 90 mins. Water added. In C.C. In mols. 25 820 50 1,280 100 2,200 200 4,040 300 5,880 400 7,720 500 9,560 600 11,400 $00 13,240 800 15,080 Mean percentage of BaC1, decomposed.Numbers actually ohtnin d. 98.2 98.2 96.3 96.8 96.3 94.0 95-5 92.5 88.2 91-2 85.2 80.5 81.4 79.7 67.0 63.9 65-1 49-9 47.4 52.4 31.5 34.2 32.2 33.7 31.8 24.8 - 24.8 10.3 10.3 10.8 9.7 These results are graphically represented in Curve n Plate 11. 10. Curve I3 (Plate 11) seems to us to show tliat with a low tem-yerature and a short time (30 minutes) the process of chemical change is retarded to a proportioiiately greater extent by a large, than by n small quantity of water of dilution. Curve C which ex-hibits the inffuencle of the water of dilution a t a low temperatuw but when the change is alloned to proceed for a longer time (90 minutes) shows the same ;rregularity in the action of the diluting water. And from Curve D m7e conclude that when the change is allowed t o proceed for a considerable time (90 minutes) the retarding influence exerted by large masses of dilating water is proportionately greater than that exerted by small masses even whan the temperature is allowed to rise to 1 6 O or 18".11. In order to explain these results we advance the hypothesis that under the conditions of experiment a number of hydrates of barium chloride tend to form in the solution. Among these hydrates we re-gard the cryohydrate as occupying an important place. At moderate temperatures (18") and short times (30 minutes) we suppose that the tendency to formation of cr-j-o1i-j-drate is very small ; lienco the regular action of the \water of dilution (Curve A). ,4t low ternperatnres however even if the time be short the cryohydrate (alonq witjh other hydrates doubtless) tends to form and being pro-duced in presence of a large mass of one of the products of its own dissociation this cryohydrate is somewhat stable.The greater the inass of water the greater the stability of the cryohpdrate other con-ditioiis being constant and any secondary action which may be exerted by the water being disregarded. When we make the time of action I O I I ~ C T tho tcnde~cy to formatio P E R C E N T A G E O F C H E M I C A L C H A N G E Harndon & Sons Lith 3 hlarhns Lane T+'- ON THE COURSE OF CERTAIN CHEMICAL CHANGES. 65 of cryohydrate might be expected to become greater and therefore the Curve C might be expected to approach to or even to overlap the Curve B.But long time of action increases the chance of molecular collisions and hence of molecular decompositions. Curve C keeps always above Curve B but exhibits the same general form. But that time does tend to production of cryohydrate (by our hypo-thesis) even when temperature rather tends to dissociation of the same provided the presence of a large quantity of the liquid product of such dissociation be insured is shown by Curve D which exhibits much the same general form as B and C. Now the association of water molecules to the molecules of barium chloride must be attended with loss of energy from the entire system, hence the amount of chemical change becomes proportionately less as the velocity of formation of these complex niolecules (by hypothesis) increases.12. In the next series of experiments the condition of time was arranged so that the precipitated barium oxalate should have com-pletely settled down leaving a clear supernatant liquid before the close of the action A portion of the clear liquid was drawn off by means of a pipette and aspirating arrangement and the quantity of undecomposed oxalate was determined therein without previous fil-tration. TABLE VI.-Pipetting. Details as before. Time = 90 mins. Temp. = 3". Water added. In C.C. In rnols. 25 820 50 1,280 200 4,040 300 5,880 400 7,720 500 9,560 600 11,400 700 13,240 800 15,080 100 2,200 Mean percentage of BaCl decorn-posed. 98.1 97.5 91.6 82.0 67.8 58.0 41.8 27.6 15.0 5.0 Numbers actually obtained./-A \ 98.1 97.5 97.5 92.1 91.1 81.6 82-5 68.4 67.2 58.3 57.8 42.8 40.8 26.2 29.0 12.9 16.6 10.2 18.1 17.2 4.5 5.5 Curve A Plate III represents these results in graphic form. 13. A series of experiments similar to those just detailed was carried out at a higher temperature. VOL. XYXVlI. 66 MUIR AND SLATER INFLUENCE OF WATER TABLE VIL-Pipetting. Details as before. Time = 90 mins. Temp. = 16-18”. Water added. In C.C. In mols. 25 50 100 200 300 400 500 600 700 800 820 1,280 2,200 4,040 5,880 7,720 9,560 11,400 13,2# 15,080 Mean percentage of BaC1 decom-posed. 98.2 96.4 93.5 88.4 80.0 65.9 44.3 31-2 -20-6 4.5 Numbers actually obtained. ‘98.2 96.3 94.5 90.9 80.0 68.2 42.8 31.4 18.6 40 7 - - A 96.8 96.0 -92.5 - -85.9 - -80.0 - -63.7 - -45.8 - -31.8 30.6 31.0 18.9 24.4 -4.0 5.5 -See Curve B Plate 111.14. These results are we think in keeping with the hypothesis of par. 11. Curve A shows it is true an almost regular retarding influence exerted by the water of dilution but we believe that the “pipetting method” of estimation is more delicate than the “ filtration method,” and that the inflnonce of formation of cryohydrate &c. is shown by the former method in the early as well as in the later portions of the curve. Curve B where the temperature is higher and therefore. where one of the conditions of formation of cryohydrate is not fulfilled, exhibits the special influence of water of dilution under consideration only when the quantity of water becomes somewhat large.The same general deduction may be made from the analogous filtration Curve D, Plate 11. 15. I n repeating some determinations in succeeding portions of the same liquid filtered from barium oxalate we were astonished to find great discrepancies between the results. If 50 C.C. were withdrawn from the filtrate immediately that or approximately that quantity of liquid had passed through the filter the quantity of undecomposed oxalate therein was found to be considerably greater than the quantity contained in the succeeding 50 C.C. withdrawn from the filtrate filtra-tion being continued with little or no intermission. We therefore made a series of determinations of the influence of water of dilution on the change under consideration the data being obtained from estimation of undecomposed oxalate contained in the second 50 C.C.of liquid which passed through the filter ' 3 0 N V H 3 l V 3 l W 3 H 3 d 0 3 3 V I N 3 3 t l 3 PLATE v. ( T~~~ mrr). BaCl and K2CZ04. 1 r l molc?rulea. IHrrison k Sons. Lith. S Martins Lane.W. C ON THE COURSE OF CERTAIK CHEMICAL CHANGES. 67 TABLE VIII.-Xeco.nd Filtrates. Series A. Details as before. Water added. Time = 90 rnins. Temp. = 3". Mean percentage of BaClz I n C.C. I n mols. decomposed. Numbers actually obtained. 190 2,200 93.4 93.7 93.1 300 5,880 71.6 70.0 73.2 500 9,560 52.6 50.8 54-4 ti00 11,400 3 4.9 32.5 37.3 700 13,240 34.2 32.5 36.0 Series B. Time = 90 mins. Temp. = 18". 100 2,200 9.5.4 96.0 94.8 200 4,040 890 91.5 86.5 300 5,8f30 80.8 81-9 79-7 400 7,320 67.4 69.9 65.0 500 9,560 51.1 53.8 48.4 600 11,400 35.3 36-0 34.7 700 13,240 32.5 35.7 32.3 31.0 30.4 In Plates IV and V these results are graphically represented by Curves A and A' respectively.The Curves B and C of Plate IV and B' and C' of Plate V represent the results of " pipetting " and- " first filtrate " determinations cnder the same conditions of time and tem-perature as those under which the data of A and A' were deter-mined. 16. From the curves of Plate IV it is apparent that the influence exerted by water of dilution on the chemical change under considera-tion a t low temperatures and moderate degrees of dilution is much the same whether the data be obtained by analyses of the liquids removed from above the precipitated barium oxalate by pipetting, o r by filtration.When however large amounts of water are added, the results show considerable differences according as the pipetting or filtration process is adopted the latter process exhibiting the change as proceeding to a greater extent than the former. We think that these results may be explained by supposing that a t a low temperature and with much water the tendency to formation of cryohydrate reaches a maximum and that under these conditions the whole system of chemical molecules is thrown into a state of strain from which it is partly relieved by the process of filtration the result of this relief being that the molecular equilibrium is upset and that the velocity of the decomposition of barium chloride by potas-sium oxalate is suddenly increased.If there be elements of truth in this explanation we should expect to find smaller differences in the results obtained by the methods-F 68 MUIR AND SLATER INE'LUEKCE OF WATER pipetting and fi Itration-when the condit,ions of experiment were rendered less favourable to formation of cryohydrate molecules. One less favourable condition is maintenance of higher temperature. The curves of Plate V (when compared with those of Plate IT) show that the results obtained from " pipetted liquid," " first filtrate," and " second filtrate " respectively agree much more closely when the temperature is 18" than when the terr,perature is 3'. Again with small quantities of water of dilution.i.e. with a condition unfavourable to formation of cryohydrate the differences between t'he results obtained by the " pipetting," and those by the '' first filtrate " methods are reduced almost to zero even a t 3" ; at this temperature " pipetted " results differ slightly but only sliqhtly from " second filtrate " results while a t 18" the results obtained by both methods are almost identical. A small number of experiments carried out at a higher temperature 50" showed that identical results are obtained a t this temperature by the three methods with a dilution of 400 C.C. and 800 C.C. It is difficult to foretell what influence would be exerted on the general stability of the system a t a lorn temperature by variations in the time of action. The numbers obtained by us seem to show that the system is in a state of greater strain a t 3" after the expiry of 30 minutes than when 90 minutes have elapsed.Thus the mean percentages of barium chloride decomposed a t 3" after 30 minutes' action as measured by analysis of the first and second 50 C.C. of filtrate were respectively :-C . C . water added. First 50 C.C. Second 50 C . C . Difference. 100 74.8 77-3 2.5 300 61.2 71.4 10-2 400 45.8 68.4 22.6 The results obtained under the same conditions save making time = 90 minutes are stated for sake of comparison. C . C . water. First 50 C.C. Second 50 C.C. Difference. 100 '32.0 93.4 1.4 300 71.5 71.6 0.1 400 63.7 65.0 1.3 The system would thus appear to be in a state of maximum strain when the conditions favourable for formation of cryohydrate are ensured viz.low temperature and much water and when the time of action is short. We are almost inclined to believe that under these conditions the cryohydrate molecules are in process of formation, whilst after a longer time the molecules are to a great extent formed ON THE COURSE OF CERTAIK CIIEMICAL CHANGES. 69 a,nd that therefore the system although yet very unstable is neverthe-less more stable than it was under the former time conditions. 17. We do not attempt to propound any exact hypothesis as to the action of the filter in inducing a sudden increase in the velocity of the chemical change. It may be that the reacting molecules are brought into closer contact i n the pores of the filter,* or it may be that when a portion of the liquid the constituents of which are by our hypothesis in a state of strain is removed from the main liquid decomposition is induced in the separated portion by the small change of temperature which is undergone by the solution during its passage through the filter or perhaps by tlie mechanical action of the surface of paper to which the solution is for a time exposed.A few determinations were niade of the amount of change when calculated from data obtained by analyses of the third 50 C.C. of filtrate. These determinations were almost identical with those obtained from the second 50 c.c. but considerably higher than those obtained from the first 50 C.C. of filtrate. The results obtained by the pipetting and filtration methods are stated in percentages of total barium chloride decomposed the rapid change which occurs while filtering under certain conditions of dilu-tion and temperature occurs in that small isolated portion of the general chemical system which is placed upon the filter but the amount of change is calculated as if that change proceeded to the same extent within the whole liquid.The exposure of the whole liquid to a veyy slightly higher temperature or to passage through a filter would not necessarily produce a change to the same extent as is produced in the small isolati d portion. Certain experiments which we have not as yet continued seemed to show tliat if when strontium sulphate is precipitated from a dilute solution of the chloride by addition of sulphuric acid and when tile liquid abore the precipitate is perfectly clear a portion of this clear liquid be isolated from the main portion by wit,hdrawal in a pipette the clear liquid so isolated quickly becomes turbid because of r en e w c d pr e ci p i t a t i o n of strontium s ul p h ate .18. The individual results contained in many of the foregoing tables shorn7 considerable discrepancies among themselves ; these discrepan-cies art. not however in our opinion contradictory of the hypothesis which we have zdvanced. The method of determining the amount of chemical change which we have used is fitted to give fairly accurate results if the mean of several determinations be adopted but the occurrence of small differ-ences between the numbers actually obtained is to be expected. The discrepancies are most marked when dealing with dilute solu-* In connection with this compare Baylej this Jomnal 1878 Trans.p. 304 '70 MUIR AND SLATER INFLUENCE OF WATER tions a low temperature and a short time of action ; they become leas marked when the temperature is high the solutions less dilute or the time longer ; further the results obtained by the " pipetting " method show on the whole fewer discrepancies than those obtained by the method of " filtration." In other words when the system is by hypo-thesis in a condition of strain differences between the individual results are noticeable but when the conditions favourable to production of such strain are removed the discrepancies among the individual results tend also to disappear. 19. In paragraph 4 it was stated that the influence of water of dilution on the reaction between strontium chloride and sulphuric acid had been examined but that the numbers obtained led apparently to no general results.We think that the great discrepancies noticed between the numbers obtained may be explained by the hypothesis already advanced. Sulphuric acid forms many hydrates with water and is doubtless able to attach to itself large numbers of water molecules. If our hypcthesis be true we should expect a system containing sul-phuric acid much water and such a salt as strontium chloride to be in an eminently strained condition and therefore to be capable of having its equilibrium upset by very small amounts of impressed force. Mow unless very special precautions were taken to maintain the conditions of each experiment altogether unchanged during the whole course of that experiment and unless a very delicate method of deter-mining the amount of change under given condit,ions were adopted, we should expect to obtain discordant results.But we did not adopt such speciaZ precaution nor was the method of measuring the amount of change characterised by extreme delicacy.* The results obtained were therefore discordant as indeed we shouId expect them to be if our hypothesis be correct. We subjoin a few of the actual numbers obtained. * The method consisted in filtering a portion of the clear liquid precipitating undecornposed strontium chloride as carbonate dissolving in standard acid and determining residual acid by titration with standard alkali PLATE VI.( TABLE. X). BaZNO.pnd &C,0,=1:lmoleoules ON THE COURSE OF CERTAIN CHEMICAL CHANGES. 71 TABLE TX.-Strontium Chloride and Xulphuric Acid. 20 C.C. SrC1 solution used = 354.64 mgrms. SrCI,. 878 C.C. H,SOk solution used = 877.52 mgrms. H2S0,. SrC12 H,SO = 1 4 mols. Temp. = 16-18". Percentage of SrC12 decomposed. h C.C. water added. %me = 60 mins. Time = 18 hour; r - 25 50 75 100 125 150 175 200 225 250 759 - -73.6 -62.9 66.1 77.3 52.8 67.2 64.7 47-7 - -33.7 49.6 -46.2 -27.4 -29.7 - -27.8 - ----Time = 60 mins. SrC1 €€,SOa = 1 2.5 mols. 25 70.1 67.5 85.5 -50 745 60.4 56.4 -90 52.0 53.0 -100 49.5 51-7 55.1 -125 45.2 51.1 39.4 50.2 150 36.9 2'2.5 21.5 32.2 175 14.2 24.8 10.6 -20.As a further test of the value of our hypothesis we carried out a short series of determinations of the influence of water of dilution on the change which occurs when barium nitrate is decomposed by meam of potassium oxztlate. TABLE X.- Barium Nitrate and Potassium Oxalate. Used 10 C.C. Ba2N03 solution = 800 mgrms. Ba2N03 and 9.4 C.C. KzC204 solution = 509 mgrms. K2C20a. Ba2N03 K2C204 = 1 1 mols. Pipetting. Time = 90 mins. Percentage Bs2N0 decomposed. -L C.C. water added. 3". 18". ' 100 85.4 91-8 300 38.9 76.4 500 39.4 50.3 700 9.1 25. 72 MUIR AND SLATER INFLUENCE OF WATER Filtration (first 50 c.c.) results. Time = 90 mins. Temp. = 3". C.C. water added. 0 25 50 100 300 500 700 Percentage Ba2N0 decomposed. 98.7 96.9 95.8 83-0 70.5 53.0 29.5 These results are represented graphically in Curves A B and C, Plate VI.The " pipetted '' results show little or no indication of formation of hydrates in the " filtered " results on the other hand such indication is clearly shown when the quantity of water of dilution becomes con-siderable. 21. We subjoin certain data taken from Guthrie's papers regarding the cryohydrates of the salts employed by us. BaCI2.2H,O cryohydrate solidifies a t - 8" with 37 mols. of water. SrC12.6H20 37 9 9 -17" 7 ) 23 > 7 CaC12.3HzO 9 7 9 7 -37" , 12 7 7 Ba2N03 Y 7 7 7 - 0.8" , 259 7 7 K C 2 0 4 7 7 7 7 + 6.3" , 17 9 7 I n the reaction in which calcium chloride was decomposed by addition of sodium carbonate the influence of water of dilution on the change was regular; because on our hypothesis the tendency to formation of the cryohydrate of calcium chloride is but feebly marked under any ex-perimental conditions realised by us.I n the reaction between barium nitrate and potassium oxalate the influence of water of dilution is regular ; because the cryohydrate of barium nitrate tends to be formed in considerable quantity under the experimental conditions employed. B u t if the determinations of the amount of change are made in the filtered liquid then the results show large differences from those obtained by the " pipetting " method ; because the equilibrium of the system although sufficiently stable to resist overthrow by the mere process of pipetting off a portion thereof is nevertheless disturbed by the process of filtration.Further the influence of water of dilution on the change which occurs when barium chloride and potassium oxalate mutually react, is irregular ; because-on our hypothesis-the cryoliydrate of barium chloride tends to be formed under the conditions of experiment but not to be formed to such an extent as renders the system compara-tively stable Hardson k Sons.Lith. s' Martins Lane.W ON THE COURSE O F CERTAIN CHENICBL CHANGES. 73 A comparison of the " filtration " and " pipetted " curves for barium nitrate with those for barium chloride under the same conditions of temperature and time (A and C Plate TI with C and B Plate IT), shows that the difference between the results obtained by the two methods is much greater in the case of the nitrate than in that of the chloride.On our view of the influence of water of dilution the very stability of the barium nitrate solution (because of the large formation of cryo-hydrate) would be looked on as the reason why when that stability is overthrown the velocity of the chemical change is so largely in-creased. 22. But there is an aspect of the influence of water of dilution other than that which we have hitherto regarded. The addition of much water might be supposed to bestow upon the reacting system a greater degree of molecular mobilit,y than would be possessed by a more concentrated solution. Chemical change should therefore occur with greater readiness in the former than in the latter liquid. Rut a t the same time the chances of molecnlnr encounters, and therefore of molecular decompositions occurring in unit time must be smaller in dilute than in concentrated liquids notwithstanding the greater niobility of the former sjstcm.Therefore we should conclude that the influence on the chemical change of small impressed forces would be more marked in dilute than in less dilute solutions while a t the same time the total amount of change under the same conditions would be greater in the more concentrated solutions. Deville (PM. Mnij. [4] 32 365) has advanced a theory of the influence of dilution in which he supposes that energy of position is a,ctually gained by the reacting molecules when dilution is increased, and that this energy may be changed into mechanical work wherebr again heat may be evolved sufficient t o raise some of the chemically active molecules present to their dissociation- temperature i e .to sliatter them into their constituent atoms. But without pushing the influence of dilution so far as this we think we are justified in supposing that in addition to the loss of energy involved in the formation of complex (cryohydrate) molecules there would be a gain of mobility in the case'of those molecules of the reacting bodies which had not thus associated to themselves large num-bers of water molecules. Dilution would thus exert two more or less opposing actions. The general result of a series of experiments carried out by one of XIS,* is in keeping with the hypothetical deduction that impressed force should produce greater diflerences in the amount of chemical change when dilute than when more Concentrated solutions are used.* See nest paper 74 MUIR AND SLATER IXFLUESCE OF WATER 23. How should t,ime influence the process of chemical change, specially studied by us when taking place in dilute solutions as compured with the influence OF the same variable on the same change occurring in more concentrated solutions ? The influence exerted by time we should expect to vary according as a low or a high temperature is maint,ained. At a low temperature formation of cryohydrate would tend to be increased by increasing time of action ; but if dilution act by increasing molecular mobility as well as by increasing formation of complex (cryohydrate) molecules we should almost expect that the differences between the total amounts of change, in long and in short times would be less in dilute than in concen-trated solutions.In a tolerably concentrated solution a t a low temperature we have, by hypotlhesis a tendency to formation of cryohydrate molecules ; water is added and thereby the tendency aforesaid is increased but simultaneously molecular mobility is imparted to the system ; the greater the amount of mobility the more rapidly will the chcmical change proceed in other words in dilute solutions the cheniical change will complete itself more rapidly than in less dilute solutions. Now if the tendency to formation of cryohydrate be small e.g. if temperature be somewhat high tjhe influence of time shonld probably be less marked than if the tendency to formation of crgohydrate be large.The results already detailed are arranged in graphic form in Plate VII with the view of illustrating the influence of time. Curves A and B which represent results obtained a t So show little difference between the action of time in concentrated and in dilute solutlions ; but Curves C and D show that a t a higher temperature ( l 8 O ) the influence of time is less marked when much water of dilution is present. If however the difference between the times of action were made very considerable the temperature being somewhat higher than that at which cryohydrate may be supposed to form rapidly then we think that the curves representing the results a t the two times should diverge largely. Now the curves of Plate VIII representing results after 90 minutes (A) and after 5 hours (B) at 18" diverge very largely.The following table contains the results obtained for 5 hours' action at 18" by the pipettiug met'hod (Curve B of Plate VIII) :-It is not easy to answer this question on & priori grounds PLATE Vn. BaCI d K&O,. 1 1 molecules. - 18'. Harrison k Sons. Lith. S! Martins Lane. W. ON THE COURSE OF CERTAIN CIIEXICAL CHASGES. 75 C.C. water added. 50 100 300 500 700 900 1100 Mean percentage of BaC1, decomposed. 97.6 95.6 85.2 75.7 59.3 44.7 18.7 Numbers actually obtained. 95.6 93.6 85.5 84.9 77.4 74.1 60.4 58.1 441.7 -18.3 19-1 24. Every chemical system appears to tend towards that condition of equilibrium the attainment of which is marked by the greatest loss of energy.It may perhaps be more correct to say that the entropy of chemical systems &.,-that portion of the energy which, being dependent on the configuration of the parts of the system is available as mechanical energy,-continually tends to become less. This tendency t o dissipation of energy may be arrested in various ways among others by impressing upon the system what may perhaps be described as an artificial state of equilibriuni. Thus the condition of most stable equilibrium for a system origi-nally consisting of barium chloride and potassium oxalate molecules, would be that in which barium oxalate and potassium chloride mole-cules are produced but by adding much water a portion of thereacting molecules are by our hypothesis loaded with water of hydration.In this loading energy is lost, but less energy than would be lost by the formation of molecules of potassium chloride and barium oxalate ; the system is therefore in an unstable condition but a certain degree of stability is impressed upon it by the presence of a large mass of one of the products of dissociation of the complex and unsta'ble hydrated molecules. We have thus a system in a state of strain because of the stress between its parts. A small force may be sufEicient to relieve the strain, and this relief may be attended with a rapid rearrangement of the p a t s of the system and with a large decrease in the entropy of the system. * 25. If the state of the system which we have endeavoured t o picture in the foregoing sentences be at all correspondent with the actual state of the system we should expect small differences in physical conditions to produce much greater variations in the amount of chemical change in dilute than in more concentrated solutions.The following results are in keeping with this supposition :-* We seem to have here an action opposed to that designated '' chemical induction," by Bunsen and Roscoe (Phil. Trans. 185'7). The state of a system of heterogeneous iiiolecules capable of mutual chemical action is compared by these chemists to tha 76 MTJIR AND SLATER INFLUENCE OF WATER TABLE XI.-Pipetting. BaCl K,C,O = 1 1 mol. Temp. = 20". Water of dilu-tion = 1100 C.C. Percentage of BaC1 decomposed. 7 -,\ Time = 5 hours. Diff. Tinie = 28 hours. DifT. Roughened beaker .. 31.2 34.8 3.8 67.8 66.3 1.5 Broken glass in beaker . . . . . . . . . . 9.7 39.8 30.1 66.3 60.6 5.7 Oxalate added with constant stirring. . 11.0 37.0 26.0 67-1 59.9 '7.2 Oxalate added on sur-face; after 15 minutes whole stirred up. . 14.0 39.1 2.5.1 62.7 62.1 0.6 Temp. = 20". Time = 90 mins. Percentage of BaClz decomposed. Dilution = 800 C.C. Diff. Dilution = 700c.c. Diff? r A rConghened beaker . . 1.5 3.5 2.0 24.3 21.7 2% Broken glass in beaker . . . . . . . . . . 1.0 3.0 2.0 7.0 11.0 4.0 Oxalate added with constant stirring. . 6.8 3.8 3.0 22.4 16.5 5.0 Oxalate added on sur-face; after 1 5 minutes whole stirred up. . . 3.0 1.0 2.0 11.1 11.1 0.0 Temp. = 20". Water of dilution = 700 C.C. Percentage of 13aCl2 decomposed. 'l'iine = 90 mins.Diff. Time = 24 hours. Diff. /< / . - Roughened beaker . . 24.3 21.7 2% 81.6 -Broken glass in beaker . . . . . . . . . . 7.0 11.0 4.0 7'3.3 -Oxalate added with constant stirring . 22.5 16.5 5.9 $9-3 81.1 1.8 Oxalate added on sur-face; after 15 minutes -- whole stirred u p . . 11.1 11.1 0.0 79.3 -of a wire with a certain weight attached; induction tcnds to rtxuove the weiglit. But the addition of water to the system considered by us tends to increase this weight and so to elongate the wire. Reniow the weight however and the wire suddenly contracts cc OW TIIE COURSE OF CERTAIN CHEBIICAL CHANGES. r l Time = 30 mins. Percentage of BaC12 decomposed. .--./- \ Temp. = 20'. Temp. = SO". Dilution = 800 c.c . 0.0 1.0 , = 700 C.C .3.7 s3.7 (See also curves of Plate VII.j 26. From these results it is apparent that physical differences affect the total amount of chemical change when much water of dilution is pre-sent and the time of action is moderately short (see 700 C.C. 90 minutes). Further that if the amount of dilution-water is made very large the differences between the results obtained become imuiense showing how the action is disturbed by very small alterations i n physical conditions (see 1100 C.C. 5 hours). But that if the time of action be prolonged, the results become tolerably concordant and that those physical con-ditions which were made the subject of examination but slightly affect the final amount of chemical change when that cliange is allowed to proceed f o r prolonged periods even when in exceedingly dilute liquids (see 1100 C.C.28 hours and 700 C.C. 24 1 ~ 0 ~ s ) . Further we conclude from these numbers that if the amount of water of dilution be so large and the time of action be such that but a very small amount of chemical change ensues then the amount of that change is only slightly affected by the changes in physical conditions to which the liquids were subjected in our experiments (see 800 C . C . 90 nz inutes) . The results of Table XI bearing upon the influence of temperature corroborate those of par. 22 viz. that in dilute solutions rise of teni-peratnre largely increases the amount of chemical change the experi-ments made with 800 C.C. of water of dilution however show that if the amount of change a t a moderate temperature be almost nothiiig a rise of temperature scarcely affects the total decomposition. 27. We think that the experiments recorded in the present paper justify our general hypothesis viz. that the amount of chemical change which occurs when barium chloride and potassium oxalate are mixed in the proportion of 1 1 molecules is irregularly affected by variations in the mass of water of dilntion present because the entire system is brought into a state of strain due to the stress between its parts ; and that the principal forces of which this stress is compounded, are the force tending to produce cryohydrate-and other hydrated-molecules the force tending to split up these molecules and the force tending to separate and so to impart greater mobility to the chemi-cally active molecules of the system. The results which we have obtained are it seems to us corrobora 78 MUIR ON THE INFLUESCE OF TEMPERATURE tive of what may be called " the kinetic theory of chemical action," viz. that systems apparenLly stable are continually undergoing atomic interchanges. We have spoken of a system of chemically active molecules as being, under cerhin conditions in a state of strain ; we believe that analogies with this state may be found in supersaturat'ed solutions,* colloidal molecules,? a,nd that " particular condition of bodies in which they are the dkbris of some compound and not proper chemical compounds of their constituents."