VI.-Action of the Halogens o n t h e Salts of Trimethylsul~7LiiLe. By LEOKARD DQBBIN, Ph.D., and ORME MASSON, M.A., D.Sc., Chemical Laboratory of the University OE Edinburgh. I. Action qf Iodine on the Haloid Salts. WHEN dry trimethylsulphine ioaide is shaken up with an ethered solution of iodine, i t becomes converted into a heavy black tarry liquid, which smells strongly of iodine; this is easily soluble in alcohol, and it reappears unchanged when the solution is evaporated. It decomposes slowly when warmed with water, iodine being evolved. It was not found possible to obtain the substance in a state fit for analysis, but there is no doubt that it is of the order of the so-called polyiodides. Jiirgensen ( J . pr. Chenz., 3, 338) obtained a similar compound of triethylsulphine, which gave a crystalline double salt with mercuric iodide.The bromide and the chloride of trimethylsulphine behave in a similar manner ; but, if the tarry product is washed with ether and dissolved in alcohol, large reddish-black crystals arc obtained when the solution is evaporated over sulphuric acid. These crystals smell of iodine and are not stable in the air. I n thtse circumstances, it was not thought worth while to analyse them, but their general resemblance t>o the more stable compounds described below leaves no doubt in our minds that they consist respectively of the bromo-diiodide and the ch loro-&iodide of trimethylsulphine,-Me,SBrI, and Me,SClI,. A compound in all respects similar to, and probably identical with, the second of these, is obtained by acting on trimeth~lsulphine iodide with iodine monochloride.11. Actio?& of Bromine o n the Iodide. When bromine vltpour is poured over dry trimethylsulphine iodide, niuch beat is developed and the colonrless crystals melt to a dark red liquid. The reaction goes more quickly, and, in fact, with violence, if the bromine is added in the liquid state. When addition of bromine produces no further action, the heavy red oil is exposed for. a short time to the air, whereby it loses most of the excess of bromine and becomes solid. The ma,ss is then broken up, shaken two or three times with a little ether, to remove the last traces of bromine, and dissolved in sufficient hot alcohol ; as the solution cools, an abundant crop of brilliant orange-red crystals is deposited. The mother-liquorDOBBIX AND MASSOPU': ACTION O F THE HALOGENS, ETC. 57 gives a further but smaller yield when allowed to evaporate in a vacuum; these crystals are of a somewhat darker shade in colour than the others, although of the same composition.The total yield approaches rery nearly to that indicated by theory. Analysis shows the compound t o be trimethy ZsuZphine dibronziodide, Me,STBr2. The sulphur was estimated by combustion with a .niixture of five parts of potassium carbonate t o one of potassium chlorate,- a method we would recommend for those cases where the sulphur is combined directly with methyl, &c. The halogens were estimated in solutions obtained by decomposing the salt with dilute sulphurous acid: in one, the bromine was determined after removing the iodine with cup& sulphate ; in the other, the total halogen was determined as silver salt.I. 0.2625 gram gave 0.0946 gram GO, and 0.0623 gram H20. 11. 0.2564 ,, 0.0944 gram CO, and 0.0605 gram H,O. 111. 0.2329 ,, 0.1421 gram BaS04. IV. 0.2258 ,, 0.2330 gram AgBr. V. 0.1073 ,, 0.1797 gram mixed AgBr and Ag'l. 1 I Calm- I lated. C, . . . . . , . 9.87 H, , . . . . . . 2.47 s .. .. . ,. 8.79 I . . . . . * . 34.89 Er,. . . . . , . 43.96 Found. l----I-- I--- -- Properties of Trimethylsulphi?Le Dibromiodide.-The orange-red crys- tals are not deliquescent, and are perfectly stable in air. When heated, they melt a t 94-95', with partial decomposition. A brown oily distillate passes over at higher temperatures and, finally, a t about 180', methyl sulphide and iodine vapour.The crystals are somewhat soluble in cold alcohol or ether, and much more so in hot alcohol. The alcoholic solution seems to undergo slight decom- position on long standing. The reaction with water is very characteristic ; for the crystals, when shaken with it, with or with- out the application of heat, quickly melt to a black oil, which resembles the periodide in appearance. On long boiling, it dissolves in the water, iodine beiag evolved. Platinic cldoride, added t o the alcoholic solution, produces a bulky flesh-coloured precipitate.58 DOBBIN AND MASSON: ACTION OF THE HALOGENS This is soluble in hot water, from which it crystallises in the golden cubes and octahedra chai*acteristic of trimethylsiilphine platino- chloride.Estimations of the platinum confirmed its identity. I. 0.237’7 gram of the salt gave 0.0824 gram of platinum. 11. 0.2436 ,, Y, 0.0843 ,, 9 , Found. Calculated for r-A 7 (Me3S) 2PtClo. I. 11. Pt per cent. .. . . . . . . . . . 34.94 34-66 3460 Freshly precipitated silver oxide, when shaken with the alcoholic solution, is converted into halojid salt, the liquid at the same time losing its orange colour and acquiring an alkaline reaction. This is due to the formation of trimethylsulphine hydroxide. On evapora- tion, some crystals of iodoform are obtained. Silver nitrate and silver suZp3hate also decolorise the alcoholic solution, rendering it strongly acid. The corresponding salts of trimethylsulphine are probably pmduced, but the reactions have not been fully investigated.Alco- holic potash, added to the alcoholic solution till the colour is gone, gives a yellow precipitate which increases on the addition of water, and is proved to be iodoform by its smell and general properties. The filtrate from this contains halo’id salts of potassium and tri- methylsulphine. The final action of aqueous potash on the crystals themselves is the same, except that no iodoform is produced; but there seems to be a t first an unstable product formed which impaxts an opalescent green colour to the solution. Ammonia solution a t once converts the crystals of the dibromiodide into a black solid matter which, after carehl washing and drying, has the appearance and properties of iodide of nitrogen, exploding with great violence and extreme readiness.The filtrate contains haloid salts of am- monium and trimethylsulphine. I f the gas be blown from t h e mouth of a bottle of ammonia solution over the orange-red crystals, these at once acquire a dark green colour which, however, the warmth of the hand is sufficient to dispel. But if the crystals be submittfed to a current of dry ammonia, this initial green colour quickly gives place to black, which is in tarn succeeded, in the course of two hours or more, by a uniform light apple-green. These phenomena are due to the formation of a direct addition product of the composi- tion Me3SIBr2,2NH3. It is an amorphous, non-explosive, light-green solid body, stable only in an atmosphere of ammonia. In a current of dry air, it loses ammonia and reverts to the intermediate black condition, the loss being very rapid at first, then slower and slower, Ammonia gas acts quite differently.ON THE SALTS OF TRIMETBYLSULPHINE.59 though it probably becomes complete in the course of time. At a temperature of 75-80' the compound melts, ammonia is rapidly evolved, and the original dibromiodide is left. Mere solution i n alcohol effects a similar decomposition. Water dissolves it with liberation of iodine, which explains the formation of iodide of nitrogen by the action of excess of ammonia solution on the dibromiodide. The composition of this very unstable compound could be deter- mined only synthetically. It was done in the following manner. The powdered dibromiodide was placed in a porcelain boat inside a short glass-stoppered tube, the weight of the tube and boat together being known, as well as that of the substance taken.The tube was then clamped in a horizontal position, and closed with a cork through which passed two tubes-one for outlet and the other, reaching to the far end, for inlet of the ammonia. The gas, dried by passing through U-tubes containing solid potash, wars then allowed to act for a quarter of an hour, at the end of which time it was displaced for a few seconds by dry air, and, the cork having been quickly exchanged for the glass stopper, the tube was weighed. This operation was repeated several times, unt,il the weight became constant. While weighing-though this was done a,s quickly as possible-the light green colour became tarnished with black a't the edges.0.7172 gram of the dibromiodide became 0.7841 gram by absorption of ammonia. Gain per cent. 9.34 9.32 Calculated for Me,SIBr2,2NH3 . . . . Found .. . . .. .. .. .... .. .. .. .. .. 111. Action of Chlorine on t h e Iodide. Chlorine, well washed and dried, is passed into a flask containing dry powdered trirnethylsulphine iodide. This a t once melts, much heat being developed, and a pasty black substance is formed, which quickly changes to a light yellow solid body. If the flask be sur- rounded with cold water, the product will remain of a dark colour, for the reaction is checked a t the first stage ; but it n a y be completed by afterwards substituting warm water for the cold and continuing the passage of the chlorine. The final product consists of trirnethyl- suZphine dicl~loriodide, Me,SICl,.It is readily purified in the same manner as the dibromiodide, and separates from the hot alcoholic solution in an abundant crop of canary-yellow crys t'als, which closely resemble the di bromiodicle in general properties. The mother-liquor, when evaporated in a vacuum, yields a smaller crop of crystals, of a darker shade, but identical composition. On long standing, the alco-60 DOBBIN AND MASSON: ACTION OF THE HALOGENS holic solution undergoes some decomposition, distinctly more than in the case of the dibromiodide. Annlysis.-I. 0.1953 gram of the substance gave 0.0931 gram CO, and 0.0582 gram H,O. 11. 0.1847 gram gave 0.0892 gram CO, and 0.0565 gram H,O. 111. 02321 ,, 0.1959 ,, RaSOp. ~~ -- c,. ......... I .......... c12 ......... 2 .:::: :: :: -- Calculated. 13 *09 13 -05 3-27 1 3.31 11.63 - 46 -18 - 25 -83 1 - 1to.00 I - ------- Found. 11. j 111. I-- --- I - I - Pmperties.-The yellow crystals of the dichloriodide are stable in air a t ordinary temperatures. They melt a t 103-104" with partial decomposition, evolve a, brown oily distillate when further heated, and the vapours of methyl sulphide and iodine above 170". Platinic chloride, added to the alcoholic solution, gives a bulky precipitate, shown to consist of trimethylsulphine platinochloride by its general properties and by the following analyses :- I. 0.1176 gram gave 0.0404 gram of platinum. 11. 0.1621 ,, 0.2458 ,, AgCI. Calculated for ( Me,S),PtmC16. Found. Pt ............ 34.94 (I.) 34.35 C1 ............ 37-71 (TI.) 37.51 With siZver salts and with caustic potash, the dichloriodide acts in a manner precisely similar to the dibromiodide.Ammonia solution de- composes i t with formation of iodide of nitrogen. With perfectly dry amwzoizia gas, it forms the compound Me,SIC1,,2NH3, the varions stages in the reaction being hardly distinguishable from those in the formation of the corresponding dibromiodide derivative. The final product, however, is not so light in colour ; and, when exposed to the atmosphere, it not only loses ammonia but absorbs moisture, and becomes converted into a wet black paste, which is not the case with its analogue. In other respects it resembles it closely. The compo- sition was determined synthetically in the manner already described.ON THE SALTS OF TRIMETHYLSULPHINE. 61 0.2738 gram of the dichloriodide became 0.3067 gram by absorption of ammonia :- Gain per cent.Calculated for Me,SIC1,,2NH3. ....... 12.36 Found.. .......................... 12.@2 IV. Action of Iodine Monochloyide on the Cliloyide. This results, as might be expected, in the formation of trinzethyl- sulphine dichloriodide. The reaction which takes place when the iodine monochloride is poured gradually over the dry chloride closely resembles that produced by t,he act,ion of bromine on the iodide. Much heat is developed, and the salt melts to a black pasty mass. This, however, assumes a light yellow colour and dry consistency when purified by a few washings with ether. The crystals obtained by cooling the hot alcoholic solution are identical in appearance and properties with those of the dichloriodide obtained by the method already described. The yield is large.V. A c t i o n of Iodine Monocldoride on t h e Bromide. This action produces a body which, after washing with ether, separates from hot alcohol in the form of crystals, which are inter- mediate in colour between the light yellow dichloriodide and the orange-red dibromiodide, both of which it otherwise closely resembles. It melts at 87" with partial decomposition, which becomes complete a t 180-190". Combustions of the substance gave the following results, showing that it consists of t i i m e t h y h 1 p h i ? i e chlorobromiodide, ISIeJSIBrC1, I. 0.3071 gram substance gave 0.1314 gram CO, and 0.0832 gram 11. 0.3228 gram substance gave 0.1378 gram GO, arid 0.0894 gram H,O. H,O.Found. Calculated €or +- Me,SIBrCl. I. 11. C ............ 11-27 11.67 11.64 H 2.82 3.01 3.07 . . . . . . . . . . . . VI. Action of Bromine on t h e Bromide. This action resembles that of bromine on the iodide, but the pro- perties OP the products of the two reactions are quite different. Heat is evolved, and the crystals melt to a dark-red heavy liquid, which does not, solidify when washed with ether ; nor can it be crystallised62 DOBBIN AND MASSON: ACTION OF THE HALOGENS from alcohol, as it separates unchanged in appearance when the soh- tion is evaporated. When surrounded with a freezing mixture of snow and salt, it becomes more viscous but does not solidify. It smells OE bromine; and it is slowly decomposed on exposure to the atmosphere, drops of a solution of trimethylsulphine bromide (which is deliquescent) making their appearance on the surface of the liquid after a day or two.It is decomposed much more quicklywhen heated on the water-bath, and tlie bromide is then obtained in the crystalline state. Even when the subst,ance is kept in a desiccator containing sul phuric acid, there is cont'inuous, though slow, loss of bromine. When shaken with water, it is slowly decoaposed and passes into solution. As such a body could not be obtained in a state fit for analysis, it was thought desirable to make some quantitative synthetic expe- riments with perfectly dry materials. The purity of the trimethyl- sulphine bromide was first ascertained by an estimation of its bromine. 0.2260 gram of the salt gave 0.2695 gram AgBr.Calculated. Found. Br ...... .... .... 50.95 50.74 In this and subsequent experiments with the bromide and chloride, which are both extremely deliquescent, special precautions were em- ployed to obtain the salt in a perfectly dry state. It was placed in a small porcelain boat inside a glass tube, the boat and tube, with it ground glass stopper, being previously weighed. The tube was then placed horizontally, without its stopper, in a larger tube surrounded with boiling water, and closed by a doubly perforated cork. The salt being thus kept a t loo", dry air was led Over its surface for from two to three hours. The tube was then quickly stoppered and weighed with its contents. By replacing the glass stopper hy a cork provided with inlet and outlet tubes, it was possible afterwards t o operate on the weighed salt with a stream of chlorine, or of air charged with bromine vapour, and to weigh the product (after clearing the tube out with dry air) without more than a momentary exposure to the atmosphere.0.2441 gram of the dry bromide was exposed for two hours to a perfectly dry current of air charged with bromine vapour. The red oil, which was rapidly formed, continued during that time to increase in volume, in depth of colour, and in mobility. The increase in weight was 1.3'208 gram, which is equivalent to a gain of from t,en to eleven atoms of bromine by one molecule of the bromide ; and there is no reason to suppose that the absorption would have stopped there, Dry air was then passed over the surface of the liquid for several hours, aud weighings were made a t intervals ; after which the tubeON THE SALTS 03' TRIMETHPLSULPHINE. 63 and its contents were left for three days in a desiccator containing solid potash and then again weighed.The following table shows the results of this experiment :- Air passed. -- - 2 hours . . . 4 ,) ... 6 ,, ... 8 ,, ... 9 ,, ... 11 ), ... After 3 d q s orer IiOH Weight. --- 1 *5649 0 *7510 0 -5483 0 -5031 0 *477'/ 0 *4689 0 -4613 } 0.4467 Weight gained. --- 1 -3208 0.5069 0 *3042 0 -2590 0 -2336 0 *2248 0 '2172 0 -2026 Gain per cent. 541 207 124 106 95 92 89 83 Mols. of Br gained by 1 mol. of Me3SBr. 5.30 2'03 1.22 1 0 9 3 0 *go 0 -87 0 -81 Now, the calculated percentage gain, for the formation of Me,SBr,, is 101.9, so that it is obvious that this compound, if formed at all, tends, on the one hand, to absorb a large excess of bromine when in an atmosphere containing it, and, on the other, t o decompose steadily when in air free from bromine.As it is impossible, however, to isclate this or any other definite compound, a11 we can assert is that bromine does combine with trimethylsulpliiiie bromide with develop- ment of heat, and that the analogy of the dibromiodide is in favour of the supposition that the tribromide is formed. VII. Action qf Bromine on the C l b r i J e . This action is precisely similar in character and appearance to tbat of bromine on the bromide. A dark-red oil is rapidly formed, which goes on absorbing bromine as long as it is exposed to air charged with the vapour of that substance, but gives it up again, a t a pro- gressively decreasing rate, in a current of dry air or i n a desiccator containing solid potash.It is unuecessary to give the details of the experiments, as they were similar to the one described in the last paragraph: they afforded no evidence in favour of the existence of the chlorodibromide, Me3SBr2Cl, or of that of any other definite compound.64 DOBBIN AKD MASSON: ACTION OF THE HALOGENS VIII. Actim of Chlorine on t h e Bromide, At the first touch of chlorine, the colourless crystals of trimethyl- sulphine bromide assume an orange tinge ; and then they gradually melt to a clear yellow viscous liquid, in which, at first, solid orange- coloured pieces may be seen floating. The final product smells of chlorine, which it evolves when exposed over solid potash or sub- jected t o a current of dry air.Wa,ter acts on it, causing the liberation of minute bubbles of chlorine. The following quantitative experiment was made. 0.2i46 gram of the dry bromide was subjected to the action of dry chloriue for one hour, at the end of which time it was found to weigh 0.4100 gram. A second hour's chlorination increased this to 0.4169 gram ; and it did not increase further. This i s equivalent to a gain of 51.8 per cent'., or 1.146 molecules of chlorine by one molecule of the bromide, the formation of Me3SBrC12 requiring a percentage gain of 45.2. The product was then allowed to stand for two days over solid potash, after which the weight was found to be 0.3873 gram, which corresponds to a gain of 41.0 per cent., or 0.907 molecule of chlorine by one molecule of the bromide.On further exposure it con- tinued to lose weight, though slowly. The experimental evidence in this case, therefore, decidedly favours the belief that trimethylsulyhine dichlorobromide, MeJSBrC12, is formed, which is what one would expect from the analogy of the dichlor- iodide. No doubt, the very different physical properties of chlorine and bromine may account for the fact that trimethylsulphine bromide (and, as will be seen, chloride) absorbs but a sniaIl excess of the former, though an apparently unlimited excess of the latter. IX. Action of C h l o h e o n the Chloride. I n order to observe this action properly, it is necessary to have both the chlorine and the chloride absolutely dry.If this precaution be taken, the crystals will turn yellow and begin to melt the instant the chlorine touches them. There is no sensible development of heat. Afler a few minutes the whole mass becomes converted into a clear pale yellow liquid, which undergoes no further change of appearance, though i t goes on absorbing chlorine f o r some few minutes longer. This product has some charact,eristic properties. When exposed to the air f o r a very brief period, i t suddenly solidifies; and when once this (apparently physical) change has taken place, the substance is not in the least affected by subsequent exposure to a chlorine current, however prolonged. Thus, in one experiment, the current of chlorine was stopped as soon as the product appeared to be uniformly liquid (five minutes from comniencement), and dry air was thenON THE SALTS OF TRIJLETHYLSULPHIXE.65 weight of chloride teken. passed through to sweep out the tube. This caused the substance to solidify, It was weighed and then again exposed to chlorine €or half an hour; but the weight was found to be unchanged, though the quantity of chlorine taken up mas not nearly so large as in other experiments. I n another case, after nearly the maximum amount had been absorbed (twenty minutes from commencement), some accidental cause produced sudden solidification ; and the rapid escape of dissolved chlorine caused the formation of a number of small crater-like excrescences on the surface. The action of a drop of water on the solid compound is particularly striking, for it causes a violent effervescence of chlorine.At the same time trimethyl- sulphine chloride passes into solution. Absolute alcohol and ether both decompose i t in a similar manner; but the cblorine is, of course, not liberated as such. The chlorinated chloride is decomposed more rapidly in dry air than any of the other compounds already described. The purity of the chloride employed in the experiments tabulated below was tested by means of a chlorine estimation. 0.3100 gram of the salt gave 0.3899 gram of AgCI. Calculated for Me,SCl. Found. C1 .... .. .. .. .. 31.55 31-1 I weight of product. In every experiment, the salt was dried for fi-om two to three hours in the manner already described. In experiment I the tube was swept out wihh a current of dry air before weighing, so that the product assumed the solid state.In the other cases, this was avoided by weighing the tube full of chlorine (with its cork, and tubes closed by caoutchouc caps) ; and the weight to be deducted for this extra chlorine was determined as accurately as possible by prelimi- nary experiments. In experiment IV it was proved that the maximum weight had been gained. per cent. No. of experiment --- I ,. .... 11 ...... 111 .. .... IV .... .. c1 gained by Me,SCl. 0 -2770 0'3110 0 *2377 0.2344 0 -3874 0'5170 0 *4068 0 *4036 39 -85 66.23 71 '14 72 *18 0 -631 1.049 1 a137 1.144 (maximum) Time allowed. -- In order to prove bejond doubt that the reaction consists in the simple addition of chlorine, and that there is no substitution of that VOII.XLVIT. I?66 DOBBIN AND MASSON: ACTION OF THE HALOGENS element for hydrogen, a com'bustion of the product of experiment I was made, the results of which may be compared with those obtained from the unacted-on chloride, I. 0.2634 gram of the unacted-on chloride gave 0.3045 gram CO, and 0.1918 gram H,O. 11. 0.2770 gram of the chloride, after it had been acted on by chlorine, gave 0.3217 gram CO, and 0-2020 gram H,O. Found. r---- 7 I. 11. Calculated Before Aft.er for Me,SCl. chlorination. chlorination. C ............ 32.00 31.72 31.67 H ............ 8.00 8-14 8-10 From these experiments, it seems fair to conclude that trim,ethyZ- sulphine trichloride. Me,SCl,, is formed-a peculiar and unstable com- pound possessed of a limited power of dissolving chlorine when placed in an atmosphere of that gas.X. Action of Halogens o n the Sulphnte. We have not yet fully investigated the reactions of trimethyl- sulphine sulpbate with the halogens; hut we have ascertained by experiments the following fact8. (1.) Iodine does not combine with it in the same way as it does with the haloid salts. (2.) Bromine combines with it to form a red solid body, which resembles the dibromiodide rather tlhan the prodncts obtained by brom ination of the bromide and chloride, but is less stable than the former. (3.) Chlorine acts on it to form a product which close17 resembles the trichloride in its general proporties, and particularly in the rapid effervescence of chlorine which occurs when water is dropped on it, We are pursuing the investigation of these reactions and of those of the halogens with other salts of trimethylsulphine.XI. Summary and Theoretical Considerations. It has been shown that all the haloid salts of trimethylsulphine combine directly wich chlorine, bromine, iodine, and iodine mono- chloride. In no case does there occur a replacement of one halogen by the other. The product of each reaction may be formulated as MesSX, (if we understand X to represent C1, Rr, or I, indiscriminately), this being proved with certainty in some cases, and fairly inferred in those where proof is wanting. In the following table, which shows the ten possible variations of the general formula, a query (?)ON THE SALTS OF TRIMETHYLSULPHINE. 67 indicates that the substance has not been proved by quantitative experiments to have the formula attributed to it.This proof might be obtained in the caees of Nos. I1 and 111; in those of I, WI, and VIII, this is rendered impossible by the readiness with which the compounds dissolve excess of the halogen, a,nd by their unstable na,ture. - NO. I I1 I11 IV V V I TI1 VIII IX X - Me,BT,? .. Me3S12Br ? . Me3S12Cl ? . Me,STBr2 . . Me3SIBrCl. Me,SICl, , . Me,SBr, ?.. Me,SBr,Cl? Me,SHrCl, . Me,SCl,.. Appearance. Black tarry matter.. .... Dark crystals .......... Dark crystals .......... Orange-red crystals .... Orange crystals ........ Yellow crystals ........ Orange-red viscous liquid Orange-red viscous liquid Yellow viscous liquid.. .. Yellow liquid (or solid). . Behaviour in dry air. Perfectly st able.. Perfedy stable. . Stable. ......... Loses Br slowly.. Loses Br slowly.. Loses CI (and Br? Loses Cl.. ...... How produced. Me,SI + I, Me3SBr + I, Me,SCl + I,, or Me,SI + IC1 R.le3~I + Br2 Me,SBr + ICl Me,SI + (312, or Me,SCl + IC1 Me,SBr + Br, Me,SCl + Br, Me,SHr + C1, Me,SCl + C1, I n this list, it will be seen that those compounds are the most stable which contain one iodine-atom ; that those which contain more than one are difficult to purify ou account of peculiarities which seem to characterisc many of the known polyiodides of oiypnic bases ; and that those which contain no iodine tend to break up into their com- ponents, except in an atmosphere of chlorine or bromine, a,s the case may be. As regards the constitution of tlhese compounds, three theories are worthy of consideration.The first would regard them as examples of mere nioleculttr combination, and would write the general formula Me,SX,X,. This, however, is not in accordance with the nature of the decomposition which the crystalline members of the series undergo when heated, nor with the fact that so high a tempera- ture is required to effect it. The second theory is represented by X the formula Me3S-X<X, and assumes that the halogen-atom of the original salt becomes triad. I n support of this may be quoted the general reactions of the dibromiodide and the dichloriodide, and particularly the fact that they combine with two iuolecules of ammonia ; also the ease with which the trichloride breaks up, inasmuch as we should certainly expect the monad group C1, to be very unstable if it exists a t all.But this theory obviously cannot account for the com- bination of trimethylsulphine sulphate with bromine and with chlorine; and this fact must be regarded as affording it strong argument against F 268 DOBBIN AND MASSON: ACTION OF THE HALOGENS, ETC. the theory ; for we have already pointed out the striking similarity between the compounds so formed and some of those derived from the haloid salts. The third theory is capable of general application to all the bodies we have described. It assumes that they are trimethyl-sulphinic salts, Me3S-X, in which the sulphur atom is hexad. /x ‘X The analogy between these bodies and the so-called polyiodides (Weltzien, Ann. Chem. Pliarm., 86, 292; 91, 36 : Tilden, this Journal, 1865, 99: Jorgensen, Ber., 2, 460; J.pr. CYhem., 3, 145, 328) and polybromides (Marquart, J. pr. Chern., 1, 429) of the nitrogeu bases is well marked. Moreover, Tilden (this Journal, 1866, 145) describes a dichloriodide of tetrethylammonium, which he obtained by acting on the chloride with iodine monochloride; and Weltzien (Ann. Cliem. Pharm., 99, 11) also obtained some similar compounds by indirect means. Professor Crum Brown suggested t.0 us, therefore, that we should investigate the hitherto untried action of bromine on the iodide of tetramethylammonium; and we have found that this reaction produces a red crystalline body so closely resembling the dibromiodide of trimethylsulphine as to leave no doubt of its true nature. Moreover, we have found that both bromine and chlorine unite with the sulphate of tetramethylammonium to form compounds which cannot be distinguished by their general properties from those obtained from the corresponding trimethylsulphine salt. We purpose making a complete investigation of the halogen com- pounds of the tetramethylammonium salts ; but what we have already observed, taken together with the work of t,hose chemists whom we have quoted, seems to justify us in saying that there is a large class of substances formed by the union of halogens with the salts of organic bases; and that any theory, to be acceptable, must be appli- cable equally to such examples as the ‘‘ pQlyiodides,” the dibrorniodide of trimethylsulphine, the trichloride of trimethylsulphine, the product resulting from the union of bromine with tetramethylanimonium iodide, and the products obtained by the union of bromine and chlorine with the sulphates of trimethylsulphine and tetramethylammonium. If, therefore, the bodies we have described are trimethylsulphinic salts, containing hexad sulphur, it would seem that nitrogen is capable of becoming heptad. Further evidence is, however, wanted on this point. In conclusion, we desire to express our gratitude t o Professor Crum Brown for having given us the great benefit of his aid and advice in this investigation.