THE QUARTERLY JOURNAL OF THE CHEMICAL SOCIETY. X.-Pho tochemical Researches. By Professor BTJNSEN and H. E. ROSCOE,B.A. PH.D. PHOTOCHEMICAL determinations which pretend to anything more than a rough approximation are surrounded by difficulties of SO considerable a nature that up to the present time all attempts to gain a knowledge of the laws of the chemical action of light have been fruitless. In a research on this subject which has occupied us for the space of nearly two years we were induced to commence a careful exarnina- tion of the action of light on solutions of chlorine bromine and iodine as well when alone in solution as in the presence of hydro- genous organic sntxtances. Although these our first experiments were not carried on with a view to publication but merely for our own instruction in order to gain a starting-point for the solution of the question which we had proposed to ourselves we nevertheless are obliged to make known these results in consequence of the subsequent publication of a research on the same subject by Dr.Wit t wer .* The experimental and theoretical results found by Dr. Wittwer stand in such complete and inexplicable contradiction to those obtained by us that we have thought it necessary critically to repeat all his experi- ments in order if possible to arrive at the cause of this difference. As however in this respect our endeavours have not been successful we must confine ourselves to a scrupulous communication of our * Pogg. Ann. xciv. 597. VOL.VIIT.-NO. XXXI. 0 PROFESSOR BUNSEN AND DR. H. E. ROSCOE’S experiments from which alone the reader will be able to arrive at any satisfactory conclusion. Dr. With wer undertakes to measure the chemical alterations which chlorine water undergoes when exposed to the action of light. The value of his determination depends therefore chiefly upon the accuracy and reliability of the method by which the amount of free chlorine in the solution is estimated. Dr. Wittwer has used the old determi- nation of chlorine by means of a solution of arsenious acid coloured by indigo. This method which as is well known never gives very exact results especially when the solution of chlorine is very dilute has long since been replaced by much better processes.In all our experiments we have employed the much more accurate and convenient iodonietric method which even with the greatest dilution of the chlorine water gives a degree of accuracy attaiuable by very few analytical processes. The strength of the solution of iodine dissolved in pure iodide of potassium was determined by volumetric analyeis* of a known weight of pure bichromate of potash dried at about 3OO0C.,t and gave the following values :-I. 11. 111. A 0.0808 0-0725 0.1257 n 2 2 3 t 174.0 173.6 173.0 t 58.3 85.0 62.2 Found OL 0.0007138 0.0007077 0.0007043 As mean of these three experiments we obtain ~~=0~0007086 for the amount of iodine contained in one burette division. In order to obtain the value of CY still more exactly the iodine solution was com- pared with another more concentrated normal iodine solution prepared at a different time and from different materials whose mean value was found from the following experiments to be a=0*002416.I. 11. 111. A 0.1755 0-1596 0.2029 n4 3 4 t 61.1 59.2 59-0 t 58.2 7.9 22.1 Found a O*002412 0.002407 0*002428 * For a description of this method and the meaning of the symbols A n t &c. see pp. 219-236 of this volume. f-Ann. Ch. Pharm. Bd. Ixxxvi. 281. In this and in all the following experiments the air contained in the small flask was driven out before the decomposition began by the addition of a few pieces of carbonate of soda to the acid solution. PHOTOCHEMICAL RESEARCHES. To oxidise equal measures of a dilute solution of sulphurous acid 144.0 burette-divisions of the former and 42.1 of th.e latter normal solution were required.Hence the value of the a of the first solution is found to be 0*0007063 the value directly determined being 0.0007086. The mean of these two numbers-0~0007075-is that used in the greater part of the following volumetric anajyses. In like manner we have satisfied ourselves by experiments which we here omit that during the lapse of a whole year the normal iodine solution does not undergo any change which could influence the accuracy of the experiments. The dilution of the sulphurous acid used in these experiments cannot give rise to error because,according to the principle of the method employed by us a slow alteration of this acid liquid has no appreciable influence on the accuracy of the results.* Although the accuracy of the general iodometric method has been proved by very numerous experiments still it does not here appear superfluous to show by a few examples the degree of accuracy in the estimation of chlorine which can in this way be attained.0.2529 grm. pure bichromate of potash was boiled with hydrochloric acid and the liberated chlorine collected without loss in a solution of iodide of potassium. The analysis of this solution gave- n=5 t=62*0 t1=51-5 01=0*002504. This gives 0.18089 grm. of chlorine the amount of chloriue equiva- lent to the salt employed is 0*18091. We cite another experiment on the density of chlorine gas made according to the iodonietric method by Dr.Landolt. A current of pure and dry chlorine was passed through a tube drawn out at either end until all atmospheric air was expelled and then after observation of the barometer and thermometer the tube was closed by pressing two cao1Atchouc joinings at each end and opened under a solution of iodide of potassium in which the liberated iodine was estimated. The experiment gave- Capacity of the tube 31.050 cub. cent.; Temperature of the gas 22O.7 C.; Height of barometer at 25'*4 C. 0.7567". n=2 t=31*9 t1=0.3 01=0-005069. * We consiclei. this peculiarity of our method to be one of its great merita and cannot therefore agree with Mohr (,4nn. Ch. Pharm. Bd. xciii. Heft l) who abandons this improvement and for sake of convenience remodels the process accord- ing to the old volumetric method.Still less to be recommended is his proposal in the same memoir to determine the free chlorine directly without previous addition of iodide of potassium. In this case an entirely inaccurate result is obtained as is generally the case when the separation of the iodine takes place in mixed liquids instead of as the method describes in a solution of iodide of potassium. 1% PROFESSOR BUNSEN AND DR. H. E. ROSCOE'S hence 28.439 cub. cent. of chlorine at 0' and 0.76" pressure weigh 0.08997 grm. The specific gravity of chlorine as calculated from these data is found to be 2.446; that calculated from the atomic weight is 2.449. These experiments together with many others cominunicated in the course of this research prove that the detcrrnination of chlorine by the iodometric method is free from those sources of error to which the old method of Gay-Lussac is known to be subject.Entirely independent of the inaccuracies of a badly chosen method in Dr. Wittwer's experiments a source of much greater error liesin the non-consideration of the disturbing infliiences which the pheno- mena of gas-absorption and diffusion give rise to. Dr. Wittwer imagined that he had entirely eliminated these sources of error by dropping into his solution of arsenious acid a chlorine-water which contained only 1to 4 parts by weight of chlorine to 1000 of water. By the simplest experiments however he might have assured himself that by this method not even approximate results can be obtaiiied for solutions of chlorine of the strength described are influenced by the phenomena of gas-absorption and diffusion in the same way as more concentrated solutions.The amount of chlorine which is liable to be lost from solutions of the above-mentioned strength by mere dropping from one vessel to another may be seen from the following experiments :-EXPERIMENT I.-A small glass of 9,834 cub. cent. capacity fur- nished with a well-closing glass stopper was filled with chlorine- solution which was quickly emptied into a solution of iodide of potassium. The volumetric analysis gave- n =2 t= 107.3 1 =99.6 tc =0.0005952. This is equivalent to 1.94(5grm. chlorine in 1000 grms. of water. The experiment repeated with the difference that the chlorine water was slowly dropped into the solution of iodide according to the plan adopted by Dr.Wittwer in his volumetric analysis gave the follow- ing results :-n = 1 t= 107.2 t,= 1-4 01 =O.OOO5952 or in 1000 grnis. liquid 1.790 chlorine. EXPERIMENT 11.-The same volume of liquid previously employed poured quickly out gave- rh= l t= 107.3 t,= 1.3 a =0*000591i2; PHOTOCHE31ICAL RE SEARCHES. or 1.792 chlorine in 1000 water. When dropped out the amount obtained was- n=l t=107*3 t,=10*8 ct = 0.0005952; or 1.633 chlorine in 1000 parts of water. EXPERIMENT 111.-By quickly ponring out the following values Tvere obtained :-n=l t=148*3 t,=56*9 a= 0*0007075 which is equivalent to 1,838 chlorine per thousand.By dropping the same volume gave- n=1 t-148.3 t,=67-0 c~=0.0007075; or 1.635 chlorine per thousand. EXPERIMENT IV.-Quickly poured out the same volume of solution gave-n-=1 t = 148.2 t,= 62.0 tl =OmO00’i075; or 1.733 chlorine per thousand. Dropped slowly the same volurtie of liquid gave- n= l t = 148.2 t = 68.6 ct = O60Q07O75; or 1.601 chlorine per thousand. The first of the following columns shows the amount of chlorine after quickly mixing; the second after dropping slowly; and the third the loss of chlorine during the dropping expressed in per- cen tap :- I. 11. 111. 1*945 1.789 S.0 1.792 1.633 8.9 1.838 1.635 11.1 1-733 1.601 7.6 The loss of chlorine amounts therefore to a mean of 9 per cent. When vie consider that a loss of chlorine must ensue even in that portion which is quickly poured from the stopped bottle and also that a great part of Dr.TVittwer’s experiments were made with niore concentrated solutions than the above it is clearly seen that the error to which Dr. Wittwer is liable from his method of experi-menting is at least 9 per cent. of the total amount of free chlorine contained in the solution. Let us now examine the relation in which this sGurce of error stands to the accuracy of the experiments given on page 599 of the quoted memoir. In order to make this couiparison more clear we have arranged the PROFESSOR BUNSEN AND DR. H. E. ROSCOE'S following Table in which the first column shows the per-centage loss of chlorine which Dr.Wit t wer found by exposing the chlorine solu- tions to the ligbt ; and the second the loss of chlorine which according to his theory he ~hould have found if his experiments had been free from observatknal errors. The loss of chlorine produced by insolation expressed in per-centage on the amount of chlorine originally contained :-Found. Calculated. I. EXPERIMENT . 28.1 28-2 . 28-0 28-1 11 . 27.4 28.1 1) 11. . 27.3 27'3 71 . 28'1 27.5 >l . 25-3 27.3 11 m. . 41.2 41.2 97 . 42.6 41'2 71 . 40.9 41-2 11 IV. . 10'4 10'4 1) . 11.6 10.4 11 . 10.4 20'4 71 v. . 8.9 8.9 1? . 7.8 8.9 19 Y) . 8.0 8-0 vr. . 33% 33.6 1f . 35.0 33.6 19 . 35.4 33.6 1) VII. . 9.9 9-9 71 .12'4 9.9 It . 20'9 9.9 71 We have just shown that even the error occurring from diffusion I can on an average amount to 9 units of the foregoing numbers and therefore in a few of the experiments is greater than the total loss from which the theory is deduced. If under such circumstances it is difficult to understand how D r .Wit t w er could employ so inaccurate a method it is perfectly unintelligible how using this method he has arrived at results which do not in any instance show an error of more than 3 per cent. We mention another example which shows how little Dr. Wit twer has considered or removed the errors which ensue from the phenomena of the diffusion of gases. At page 608 of the memoir we read ''The action of chlorine on the vapour of water must here be noticed.In order that water be transformed into the vapoury con-dition it is well known that 1 gm. must take up 550 heat-units ; but its constituent atoms on account of this heat are less strongly attracted together; and when chlorine acts upon the vapour in Y1i OTOCHE M ICA L RE8EARCHES. presence of light the action proceeds more rapidly than with liquid water.” We find the following experiments cited as proof of this opinion :-Portions of the same chlorine-solution (of strength 3.720 pa*thou-sand) were poured into three small bottles the first was completely filled the second to one-balf of its volume the third to one quarter of i,ts volume and the three exposed to the action of the light for the same period.The first lost 17.07 the second 32.00 and the third 48.47 per cent. of chlorine. Dr. Wittwer ascribes these differences to a change of affinity arisins from the latent heat of the vapour of water ; and he has not considered that an interchange must take place between thc chlorine in solution and the atmosphere of nitrogen and oxygen which exists above the liquid; and that therefore according to the laws of gas-absorption the loss of dissolvcd chlorine is greater as the relative volume of tlie air to the watcr increases. We have repeated the experiment differing only from D r. Wit t wer’s by the circumstance that we allowed the air and chlorine-water contained in well-stoppered bottles to remain in contact during entire absence of light. After the lapse of even four hours the influence of diffusive absorption has become so visiblc as is wen from the following Table that we must conclude that D r.Wit t wer has detcrniined rather this action than that of the light. TABLEI. Chlorine in Volume. n fl 1000 water. - ljottle completely iillcd . Ilitto half filled . . . 18.13 2G.28 5 6 132.0 112-0 13:;-.2 133.0 4.717 4.341 Ilitto qnnrter filled . . Ditto ditto . . . . 11.42 10.00 2 2 12.5 129.0 3.376 3.650 The bottles were analysed four hours after filling the amount of iodine contained ill one burette-division was ct=0*0005952. The bottle which was half filled lost therefore by standing four hours in the dark 8 per cent. of its amount of chlorine ; the two tilled to a quarter of their volume lost 22.6 and 24.2 per cent.The cause of this con- siderable loss of chlorine niight be ascribed chiefly to the concentra- tion of the solution. We have therefore made a set of similar deter- niinations with dilute solutioiis of chlorine; but in this as in the former case a considerable loss of chloyine was obtained. PROFESSOR BUNSEN AND DW. H. E. I1OSCOE’S TABLE11. The solutions stood in the dark for 4hours after filling ;OL =0.0005952 t Chlorine in Volunie ji~ t t1 1000 parts. Bottle filled full . . . 18.83 1 189.0 18.8 L.303 Ditto half filled . . . 23.80 2 14.0 170.8 M48 Ditto quarter filled . . 10.20 1 180*0 106-0 1.353 i TABLE 111. The solutions stoodinthe dark for 15hours after filling OL =0.0003952. Chlorine in Volume.11 d ‘1 1000 paris. ----_I_ Bottle filled full . . 1.8.83 1 182.0 15.3 1.476 I Ditto half filled . . 29.53 2 182.0 110.3 1.377 Ditto quarter filled . . 10.60 1 181.0 106.0 1-177 Ditto ditto . . . . . 12.15 1 183% 98.0 1.174 The bottles filled to one-half lost therefore 3.66 and 6.64 per cent. of chlorine; those filled to one-quarter lost 9.98 20.2,and 20.4 per cent. respectively. If we now remember that by dropping the chlorine-water into the normal solution Dr. Wittwer must have had a still greater loss of chlorine there can be little doubt concerning the amount of reliance to be placed in the esperimental results which he has obtained. We have in our experiments eliminated these large errors ensuing from gas-absorption by insolating the chlorine-water contained in hermetically sealed tubes in which the volume of the air to that of the liquid employed was inappreciably small.Equal lengths of the same glass tube carefully cleaned and as free as possible from irre-gularities of about 18 millimetres diameter were drawn out to fine long points at each cnd. All these tubes were then dipped into a deep vessel containing a perfectly well-mixed chlorine* solution until the upper end of the tube just appeared above the surface of the liquid. The chlorine water which rises from the bottom of the vessel thus comes in contact with the air of the tube for scarcely a single * The chlorine solutions employed in all these experiments were prepared and after-wards preserved ina room from whkh all daylight was excludecl.P Irl OTOC H E NIC A L RE SEA It CH E S. second and without the slightest agitation. After a small ball of wax had been pressed on the end which rose out of the liquid the whole tube was withdrawn from the solution and by means of the blowpipe hermetically closed at both ends. A small bubble of air remained above the liquid large enough to prevent the tube from bursting by any subsequent alteration of temperature. The tube which was weighed before the experiment was after filling again weighed (by candle-light) together with the ends melted off before the blowpipe and the difference of these two weights gave the amount of chlorine-water contained in the tube. By this mode of filling the titbes the chlorine-water comes in contact for scarcely a second with a perfectly undisturbed colunin of air of only 250 square millimeters sur-face and is then irnrnediately prevented from further contact by being hermetically closed.We are thus perfectly sure that the disturbing influences of diffusion are fully eliminated. In order to avoid the same source of error during the analysis the tube was held vertically and its lower end broken under the iodide of potassium solution and afterwards the upper one also opened so that the chlorine-solution flowed into the iodide without coming in contact with the outer air. The chldrine-solution which remains on the inner surface of the tube is easily collected by washing out with a few drops of iodide-solution. The weight of chlorine (C) equivalent to the amount of liberated iodine is found according to the method described by one of us in the Ann.Chem. u. Pharm. Bd. lxxxvi. 265,* by means of the following formula c1 c= -I a(nt-t,) After these reniarlrs we proceed to a critical examination of Dr. Wittwer’s research. The starting-point and basis of all the conclu- sions contained in the memoir are found in the following sentence (p. 598) :-“By equal amounts of light the quantity of hydrochloric acid formed is proportional to the strength of the chlorine-water.” The supposition of such a proportional action appears even from a theoretical point of view in the highest degree improbable. It would require the further assumption that the chemical attraction which the chlorine exerts on the particles of hydrogen of the water is independent of the chemical attractions of the remaining bodies either already present or formed during the decomposition ; whilst daily occurring examples teach us that affinity must be considered as the resultant not only of the attractions of the combining molecules but * See also page 219 of this volume.PROFESSOR BUNSEN AND DR. H. E. ROSCOE’S 202 also of all those present within the sphere of the chemical action; and that the magnitude of the chemical affinity is changed according to the relative number and material differences of the molecules,- dependent however upon laws which are as yet quite unknown. Nitrogen in the free state is certainly one of the most indifferent bodies known and still the mere presence of this substance is sufli-cient to change the affinity of oxygen to hydrogen to such an extent that in order to induce the combination of the latter gases it is neces-sary to raise their temperature to a very different degree than that required if oxygen and hydrogen be alone present an exactly similar action is found to take place in mixtures of all gases.The numberless phenomena which we class under the term ‘‘ Catalysis,” are merely special cases in which this general action of affinity is rendered more strikingly evident. The consideration of phenomena of this kind makes it at once unlikely that the force with which chlorine decom- poses water should neither be influenced by the volume of the water present nor by the interchange which takes place between the water and the chlorine and oxygen and hydrochloric acid formed during the decomposition.Dr. Wittwer has so little coilsidered the action of these causes on chemical affinity that he states that the presence of hydrochloric acid is indzferent and specially mentions (p. 611) that he has convinced himself by various experiments of this indifference. The following facts will sufficiently show how far this statement is at variance with the truth. The gas obtained by electrolysis of pure concentrated hydrochloric acid evolved from poles of pure carbon consists as one of us has shown of exactly equal volumes of chlorine and hydrogen. When this gas freed from all trace of hydrochloric acid by passing through water is then dried over chloride of calciurn and collected with the necessary precautions t.he afinity of the two gases is such that volunies from 50 to 60 cub.cent. when exposed even on dark days to the diffuse light of a room unite with explosion. The same gas collected over tolerably concentrated hydrochloric acid may be exposed to the direct rays of the sun without any fear of explosion occurring the combination taking place gradually. Here then we have a case in which the diminution of the force of affinity of chlorine owing to the presence of hydrochloric acid is most clearly shown. It is now not difficult to prove that in the dccornposition of water by chlorine the affinity of the latter to hydrogen is altered by the formation of hydrochloric acid.Nearly two years ago we filled a glass tube of 82.3cub. cent. containing about a gramme of water with chemically pure chlorine and closed it hermetically bcforc the blowpipe. This tube PHOTOCHEMICAL RESEARCHES. -. Weig!:t of solution in grammes ..... t ........ 29.037 150.5 29378 135.7 27*.103 133'2 28'GOl 135.7 t n ........ ........ 11.4 1. 95.4 1 24'8 1 22.8 1 Chlorine in 1000 parts . Of 100 chlorine de- 1.017 037 I 0.782 0.782 composed. .... 0.0 73'4 0.0 0.0 was exposed to the direct rays of the sun and to diffused light for the space of more than twenty months without appreciably changing its colour ;and after this period it was opened under a tolerably concen- trated solution of potash. The amount of non-absorbed.gas was found to be only 2.4 per cent. of the chlorine employed. From this experiment it may be coiicluded that the affinity of chlorine to hydro- gen becomes inappreciably small as sooii as a certain amount of hydrochloric acid is formed. The following numerical data show the great retarding influence which hydrochloric acid exerts upon the decomposing force of chlorine on water :-TABLEIV. Time of Insolation 1 hour direct sunlight. Pure Chlorine Water. Chlorine Water with 10 per cent. Hydrochloric Acid. a=O*oOO7~5 Before After Before After insolation. insolation. insolation. insolation. TABLE V. Exposed for 6 hours to direct and diffuse sunlight. Pure Chlorine Water. Chlorine Water xith 10 per cent. HCI. Before After /jBefore After a=0.0007075 1 insolation.insolation. insolation. insoIation. Weight of solution taken .. 29.001 27.832 27.463 .......... 133.3 8.4 133.2 Ifl .......... 115.0 7% 24.6 n .......... 2 1 1 Chlorine in 1000 water .. 1.0338 0-00426 0.7820 Of100 chlorine decomposed . 0.0 99% 0.0 PROFESSOR BWNSEN AND DR. H. E. ROSCOE'S TABLE VI. Insolated 1hour in diffuse light and afterwards 30 minutes in direct -sunlight. Pure Chlorine Water. Chlorine Water with 3 pel. cent. Hydrochloric Acid. a=0.0007075 Before insolation. After insolation. Before insolation. After insolat ion. Weight of solutioii employed t .......... 20.831 108.8 29.48 L 107.1 28.560 108'7 tl .......... 62-0 102.0 92.0 n .......... 3 3 3 Chlorine in 1000 water .. 1.763 1.471 1 -621 01100 chloi.ine decomposed 0.0 16.1 0.0 TABLEVII. Insolated in 1hour diffuse and 30 minutes direct sunlight. Chlorine Water with 3 per Pure Chlorine Water. cent. Hydrochloric Acid. 1 1 Before After Before After a=O'OC07075 insolat.ion. insolation. insolation. insolation. Weight of clilorine water . . 29.531 27.185 31"iGL t..........108.6 108'6 107.0 t .......... 85 4 1.0 87'2 n .......... 2 1 2 Chlorine in 1000 water . . 0'8530 0.5125 0.781)3 Of 100 chlorine decomposed . 0.0 0.0 09 From Table IV. it is seen that a solution of chlorine exposed to direct sunlight during one hour lost 73.4 per cent. of chlorine. The same solution to which 10 per cent. of hydrochloric acid was added was not the least altered by exposure during the same time.According to Table V.,'a solution of chlorine exposed for upwards of six hours to direct and diffuse sunlight lost 99.6 per cent. or nearly the total amount of contained chlorine. By addition of 10 per cent. of hydrochloric acid the action during the same time is reduced to 1-3per cent. In Tables VI. and VII. chlorine water exposed to diffuse light for 1hour and to the sun's direct rays for 30 minutes lost from 16.1to 17.6 per cent. of chlorine; mixed with 3 per cent. of hydrochloric acid and subjected to a similar insolation the amount of chlorine remained unaltered. In order to give astill furthcr confirmation of thc sulJject under PHOTOCHEMICAL RESEARCHES. consideration we mention one other experiment made a year ago which shows that the decomposing force which bromine exerts upon water containing a small quantity of alcohol is considerably altered by the presence of hpdrobroniic acid.TABLEVTII. 11Ditto with ;t-to 9 per cent. Alcoholic Bromine Water. Hydrobromic Acid. II.--I. 11. 111. IT. TT. VI. 1 I Time of insolation ..... -30' -30' 45 60' Weight of solution in pimines 200 200 200 200 200 200 t ...........47.0 46'2 47.0 46.2 46.2 44.4 I,. ..........19.4 32.8 19.4 21-5 24.0 21'7 7~. (a=0'01)22503) ..... 1 1 1 1 13romine iu 1000 water . . I !)cif5 O.!j54 1-SGI 1-75! TAossof bromine per cent. .. 0-0 51.5 0.0 10.5 The lowest horizontal division in the Table shows that the loss of bromine in the liquid free from acid during an exposure of thirty minutes amounted to 51-5 pcr cent.;that the loss during the same period of the solution containing only 1per cent. of hydrobroruic acid was only 10.5 per cent. and during a period twice as long 17.75 per cent. From all these experiments thcx cannot therefore be the slightest doubt that the hydrochloric acid produced by the decompo- sition of water by chlorine exerts a very strong retarding action on the affinity of chlorine for hydrogen. We are unable to conceive what circumstances can have induced Dr. Wittwer* to ignore these plain facts as no single one of his "often repeated" experiments on this subject is given in his memoir. Dr. Wit t wer's fundamental expc:riments are however still more inexplicable.It is difficult to conceive how a mode of experiment which is accompanied by a mean error of 9 per cent. can give results whose inaccuracy scarcely reaches the error incurred by the weighing ; and how these numbers can stand in perfect harmony with a hypo- thesis which as our experiments show has no application to the decomposition of chlorine-water. Dr. W ittwer's theory starts from the supposition that aqueous solutions of chlorine lose by equal insolation an equal fraction of the originally contained chloiine,-with *Dr. Wit t wer actually proposes to employ a solution of chloride of lime decom- posed by hpdroclilo~ic acid as an absolute incnsiire of the chemical rays. PROFESSOR BUNSEN AND DR. H. E. ROSCOE'S the limitation however that the layer of chlorine water does not exmd a certain thickness.To satisfy this condition the following experiments which we have arranged in a tabular form were conducted with hermetically sealed tubes of 20 millimeters diameter and therefore smaller than the stoppered bottle employed by Dr. Wittwer. In spite of this our experiments do not show even approximation to the law founded by Dr. Wittwer. The theory requires that the per-centage decompo- sition remain constant for all concentrations of the chlorine solutions ; the results of the following Tables show that in every case a different per-centage decomposition was obtained. From Table IX. 11.94 and 15.7 per cent. 9 X. 43.0 ,,37*5 , 9 XI. 45.9 , 53.6 ,,and 404per cent. 9 XII.68.6 ,,75.3 , ,,62-8 , , xIIr. 81.8 ,,88.2 , J9 77*7 9 99 XIV. 73.3 ,,58.7 ,, I TABLEIX. Pure Chlorine-water exposed to diffuse light for 69 hours. a=0*0007075 Before After Before After insolation. insolation. insolation. insolat,ion. __I___---Weight of solution .... 32.786 27.879 29.781 28.142 t .......... 144.5 145.5 108.6 149.0 t .......... 9-0 81.0 85.4 40.0 12 .......... 2 2 2 1 Chlorine in 1000 water. .. 1.6884 1.4896 0-8752 0-7378 Of 100 chloriue decomposed . 0.0 11-94 0.0 15.7 TABLE X. Pure Chlorine-water exposed to diffuse light for 168 hours. ~~ a=O*oOo7O75 Before After Before After insolation. insolation. insolation. insolation. -~~ Weight of solution taken . . 27.199 25.686 30.B25 27.570 t .........136.8 135.2 136.2 338.2 i1.......... 25'0 1-3 3.0 GO.0 n.. ........ 2 1 1 1 Chlorine in 1000 water . . 1.8074 1.0309 0.8629 0.5394 Of 100 chlorine decomposed . 0-0 43.0 0.0 37'5 PHOTOCHEMICAL RESEARCHES. TABLEXI. Pure Chlorine-water exposed 1 hour to direct sunlight. I Before After Befom After Before After I 1 1 insolation insolation insolation ,insolation nsolation /insolation I-I-Weight of solution . 24.070 28.3 58 27*020 I 27-156 t ...... 110.0 109.8 tl ...... 5.5 30.0 n ...... 3 2 l1 Chlorine in 1000 water .... 2.669 1-4432 Of 100 chlorine de-composed ... 0.0 45.9 . . ~ TABLE XII. Pure Chlorine-water exposed for 1hour 30 minutes to direct sunlight. a=0*0007075. Before After Before After Before After insolation nsolation nsolation insolation nsolation insolation Weight of chlorine water ...24'070 27.204 27.020 27.889 27.537 28.433 t...... :I 110.0 222.7 109.9 13.4 109.8 13.2 tl...,.. 5.5 7.5 79.6 4.4 27.9 8.2 n. .. ... 3 1 2 3 1 3 ~li~orine in 1000 parts water . . 2.669 0.8574 1.0268 0.253F 0*.5880 0.2184 Of 100 chlorine de- composed ... 0.0 68.6 0.0 7 5.8 0.0 62.8 TABLE XIII. Pure Chlorine-water exposed for 2 hours to direct sunlight. I 4 1 a=0*0007075. Before After Before After Before After insolation insolation insolation insolation insolation insolation Weight of chlorine water .... 24.070 25.716 27-020 25.387 27-53?' 23'110 t....... 110'0 109.5 109.9 13.2 109.8 132 tl ...... 5.5 50.9 79.5 10.9 27.9 11.1 2 1 2 ...3 l2 &Gin; in 1000 parts water . . 2.669 0.45061 1.0268 0.1208 0.5881 0.1309 Of 100 chlorine de- 1 composed ... 0.0 81.8 0.0 88.2 0.0 77.9 PROFESSOR BUNSEN AND. DR. H. E. ROSCOE’S 208 TABLEXIV. Pure Chlorine-water exposed for 2 hours to the direct sunlight. ___-~_____-I I l a=0-0007075. Before After Before Ailer insolation. insolattion. insolation. insolation. -I-Weight of chlorine water . . ~YO~F 29~3 29.300 25.755 t ..........150.5 135.7 135-6 t .......... 11.4 88.4 1084 n .......... I 1 1 1 Chlorine in 1000 prts water. 1.01 66 0.2119 Oi 100 chlorine decomposed . 0.0 73.3 58.7 If we suppose as is most simple that the decomposing force of the light is proportional firstly to the length of exposure and to the intensity of the light and secondly to the mass of decomposing sub- stance present in the unit of volume it is seen from the foregoing Tables that the affinity of the chlorine to the hydrogen of the water first increases by diminution of the amount of contained chlorine then attains a maximum and afterwards again diminishes.Numerous instances of such a mode of action of affinity will occur to every chemist. Although after all these explanations it must appear almost unne- cessary further to examine Dr. Wit t w er’ sexperiments we have repeated the experiment cited in the research at pages 600 and 601. Our only object in continuing this examination was the wish to leave no means untried which by any chance could lead to an explanation of the above contradictions.Here also as might be expccted our attempts were fruitless. In order more fully to establish his sup- position that by equal intensities of light the amount of hydro- chloric acid produced is proportional to the contained chlorine Dr. U’ittwer shows by experiments (which again in spite of the source of error above mentioned agree precisely with the theory to be proved) that the alteration of the chlorine solution is proportional to the product of the amount of light into the strength of the chlorine- water and the time of exposure. In this experiment he determined the loss which a solution of chlorine suffers by exposure for 10 20 30 and 46 minutes each during a constant intensity of light. From S the original strength of the chlorine-water and s the strength after insolation during a time t he finds the intensity I of the light for the whole time during which his experiment lasted; and his numbers agree as far as the third decimal figure.PHOTOCH?CMTCht RESEARCHES. On page 601 of Dr. Wittwer’s memoir we find the following experiment :-t S‘ r 10’ . . . 2.1659 . . 0.11607 20’ . . . 3.9164 . . 0.11923 30’ . . . 1.7279 . . 0,11398 40’ . . . 1.5239 . . 0.11690 50’ . . . 1.3660 . . 0.115411 The first column expresses the time of insolation; the second the amount of chlorine contained in 1000 parts of liquid which before iusolation contained 2.4324chlorine in 1000parts. We have already shown that the mean error which is caused by diffusion during the dropping out of the chlorine-solution amounts to about 9 per cent.Supposing now that this error is twice as large as that actually pre-sent,-that is suppose that the real error amounts to 4.5 per cent.,- the inaccuracy to which Dr. Wittwer is then liable is so large that in his first observation for example he may just as probably have obtained the number 0.16206 instead of 0.11607 as the value of I. In order to avoid a11 disturbing influences in our experiments we conducted them precisely under the outward circumstances mentioned by Dr. W-ittwer we also chose for our insolation the noon of a cloudless day. The probability that the value of I remained constant during-our determination is four times larger than is the case with Dr.Wittwer’s experiment as we conducted the insolation according to a system which required that the light should remain constant for a period only one-fourth so long. The following is a description of the method which we employed :-Suppose that five tubes containing chlorine-water be exposed to the same amount of light during varying lengths of time; for instance the first for 18 minutes the second for twice 18 minutes the third for three times 18 minutes and so on ; all the tubes are exposed at the same moment and after the lapse of every 18 minntes each one of the tubes may be covered. As the first tube was insolated for only 18 minutes the last on the other hand for 90 minutes it is necessary that the amount of light remain constant for 90 minutes.If on the contrary all the tubes be exposed to begin with together but the first insolated for 3 minutes the second for 6 minutes the third for 9 minutes the fourth for lf2 minutes the fifth for 15 minutes and if this regular opening and covering of the tubes be repeated for each 18 minutes the light is only required to remain constant for 18 minutes and gives results as VOL. vIrr.-No. XXXI. P PlIOTOClIEMICAL RESEARCHES. accurate as the other meihod which requires a constant source of light of 90 minutes duration. In the following experiment 5 sealed tubes of chlorine-water of the strength S= 0.2641 per cent. of chlorine wcre insolated according to the preceding system so that the exposure lasted from 18 minutesup to 90 minutes. The required condition for a constant value of I was therefore given when the alteration of the intensity during every 18 minutes is taken as inappreciable ; whereas in Dr.Wittwer's expe- riment a constant intensity for l hour 10 minutes is necessary. TABLE XV. Duration of the experiment 1hour 30 minutes. Pure Chlorine-water u= 0*00059028. Length of insolation. Not 18' in 36' in 64'in 72'in 90' in insolated. sunlight. sunlight. sunlight. sunliglit,. mnlight. --. -Weight of chlorine-water . . . . 24.248 25.208 24.566 23.538 23.448 23.31 1 t . . . . . . . 125.5 125.5 126.0 126.0 126.0 126.0 f . . . . . . . 113.8 57 1 1'0 77.5 27.5 61.4 n . . . . . . . 4 3 2 2 1 1 Chlorine in 1000 parts water . . 2.6409 2.0904 1.6854 1-2214 0 GO297 0.4752 Value ofI.. . . 0.0 0.0130/ 0.01 2 0.0142 0.0186 0.01 95 ___-The lowest line of the preceding Table contains the values of I cal-culated from each experiment ; it is here seen that these numbers are anything but constant varying from 0.013 to 0.019. In reviewing the results of our experiments we find the following conclusions justified :-1. The products formed by the photochemical decomposition qf chlorine-water exert a retarding action on the amount ofth,e original aflnity of the chlorine. '2. The decomposing action of chlorine on water is therefore neither proportional to the length and intensity of the inso- lation nor to the strength of the chlorine-water. As the photochemical action is thus accompanied by a simulta-iieous change in the affinity dependent on laws altogether unknown it lvould be a completely hopcless task to eiidcavour to arrive at the DR.GLADSTONE ON THE COLOUR OF CHLORIDE OF COPPER. 211 laws of the chemical action of light by the insolation of chlorine-water. We have therefore for upwards of a year given up all attempts of this kind and have had recourse to another method by help of which we have succeeded in establishing a series of very simple relations exhibited by the chemical action of light. The simple law which governs these interesting relations we shall conimu- nicate in our next paper on this subject.