1350 PETERS : IODONIUM COMPOUNDS OF THE TYPE IR’R’’R”’ CXXXV1.-Iodoniurn Compounds of the Type 1R’R”R”’ a72cl the ConJiguration of the Iodine Atom.* By HAROLD PETERS, A.T.C. THE principal object of the work described in this paper was to try and obtain some knowledge of the arrangement in space of the valencies of the polyvalent iodine atom by determining, in the first place, whether compounds of the type IR’R’R”’ can be obtained in isomeric forms. For the preparation of such compounds (Proc., 1900, 16, 62), Prof, Kipping and I were led to study a reaction discovered by Hart- mann and Meyer (Be?*., 1894, 2’7, 502 and 1592), who bdve shown that iodonium bases of the type IR,*OH containing two identical radicles can be obtained by the action of certain metallic hydroxides on a mix- ture of iodoso- and iodoxy-compounds, in accordance with the following equation, cGH5’Io + C,H5*I02 + M*OH = (C,H,),I*OH + MIO,.As it seemed probable that by employing iodoso- and iodoxy-deriv- atives of different aromatic radicles a base of the type IR’RNmOH would be produced, we investigated the action of moist silver oxide on a mixture of molecular proportions of iodosobenzene and p-iodoxytoluene, and found that it was thus possible to prepare phenyl-p-tolyliodonium hydroxide, C,H5-I0 + C6H4Me*I0, c AgOH = $gl>I*OH + AgIO,. G 4 This, the first; known example of a mixed iodonium base,? was isolated in the form of its iodide and a number of its salts were examined. We next employed a mixture of p-iodosotoluene and iodoxybenzene, and with the aid of the same reaction prepared p-tolylphenyliodonium hydroxide, C,H,Me*IO + C,H,*IO, + AgOH = CGH4Me>I-OH + AgIO,, C,H, which again was isolated in the form of its iodide.* This communication, which was presented as a thesis by Harold Peters a t the final examination of the Institute of Chemistry, is a continuation of work already published in the Proceedings in the joint names of Professor Kipping and Harold Peters (Proc., j900, 16, 62). It was carried out at the suggestion of Prof. Kipping and with his assistance, and the cost of the materials was partly met by a grant made to him by the Government Grant Committee of the Royal Society.-F. S. K. -i- Since the publication of our original note on this work, Willgerodt and Schlosser (Bey., 33, 1900, 692) have prepared some mixed iodoiiiuiii basesiii a similar manner.AND THE CONFIGURATION OF THE IODINE ATOM.1351 This salt proved to be identical with the iodide obtained by the first method, and corresponding derivatives of these salts were also iden- tical, a fact from which it mty be concluded, provided, of course, that intramolecular change is excluded, that two of the three valencies of the iodine atom are symmetrically situated with respect to the third. For the further investigation of this problem, it was necessary to prepare a derivative containing some asymmetric group, and to then ascertain whether such a compound could be resolved into different parts. We therefore treated the phenyl-ptolyliodonium iodide with silver bromocamphorsulphonate, and thus obtained a phenyl-p-tolyliodonium bromocamphorsulphonate, which was easily isolated in well-defined crystals. A salt identical with this preparation in crystalline and other properties was obtained by interaction of the p-tolylphenyl- iodonium iodide with silver bromocamphorsulphonate, which fact affords further evidence of the position of two of the iodine valencies with respect to the third.Phenyl-p-tolyliodonium bromocamphorsulphonate was next sub- mitted to systematic fractional crystallisation ; all the crystalline deposits at first obtained were found to be identical in outward pro- perties, and in dilute solutions all had the same molecular rotation, which was practically identical with that of optically inactive salts of bromocamphorsulphonic acids. The fact that the salt is not resolved into isomerides under these conditions points to the conclusion that the three iodine valencies are arranged in one plane, but considering the large number of cases in which the use of a method such as the above fails t o resolve undoubted mixtures of optical isomerides into different fractions, further experi- ments must be made before the matter can be regarded as settled.I n fractionally crystallising various samples of phenyl-p-tolyl- iodonium bromocamphorsulphonate prepared by the first of the two methods referred to above, no difficulty was experienced in obtaining successive deposits of well-defined crystals until a very considerable proportion of the original preparation had thus been separated, after which the mother liquors always gave an oily deposit; the latter seemed to be a mixture of two substances very similar in ordinary properties, and in spite of the results of the optical examination of the first crystalline deposits, it seemed not impossible that a partial reso- lution had occurred. On examining, in a similar manner, the oily residues which were also obtained in the crystallisation of the p-tolylphenyl salt, prepared from iodosotoluene and iodoxybenzene, it was found that here, also, the last mother liquor deposited what appeared to be two different compounds.4 x 21352 PETERS: IODONIUM COMPOUNDS OF THE TYPE IR’R’”’’’ Further investigation showed that the phenyltolyliodonium bromo- camphorsulphonate prepared by the first method consisted of a mixture of the salts of phenyltolyl- and ditolyl-iodonium hydroxides? of which the latter was present in only small proportion; also that the tolyl- phenyliodonium bromocamphorsulphonate prepared by the second method contained a small quantity of what seemed to be diphenyl- iodonium bromocamphorsulphonate. The presence of the salts containing two identical radicles might be accounted for by assuming that during the preparation of the mixed iodonium bases some of the iodoso-compound undergoes decomposition into the iodoxy-derivative, 2C,H,*IO = C,H,*IO, + C,H,I, a reaction which is known to occur on heating with water, and that the latter then interacts with some of the unchanged iodoso-compound, giving the base containing two identical radicles : A different explanation altogether might, however, be possible, namely, that the substance supposed to be phenyltolyliodonium bromo- camphorsulphonate is not the salt of a mixed base, but that it is a mere mixture of diphenyl and ditolyl salts, the formation of which is brought about by an interchange of the phenyl and tolyl radicles on crystallising the bromocamphorsulphonate or a t some earlier stage in the preparation of this salt.That the former, and not the latter, con- clusion is the correct one is amply proved by the experiments which are described below and which may be summarised as follows. The iodide of the base prepared from the product of the interaction of iodoso toluene and iodoxybenzene can be separated by fractional crystallisation into a large portion of the salt having the composition of tolylphenyliodonium iodide and a vesy small quantity of ditolyl- iodonium iodide, easily identified by its characteristic crystalline form.The iodide of the base prepared from the product of the interaction of iodosobenzene andiodoxyt,oluene, does not contain any ditolyliodonium iodide but seems t o contain small quantities of diphenyliodonium iodide; the absence of the former is easily proved owing t o the fact that a mixture of ditolyl- and diphenyl-iodonium iodides is easily resolved into its components by fractional crystaliisation ; the presence of diphenyliodonium iodide, however, is difficult to prove, as it is apparently isomorphous with the mixed iodide and the two compounds are very difficult to separate ; moreover, they have both very indefinite melting or decomposing points.Salts of the mixed base, directly compared with mixtures of the corresponding diphenyl and ditolyl salts, were found to show theAND THE CONFIGURATION OF THE IODINE ATOM. 1333 behaviour of pure compounds and to differ from the artificially pre- pared mixtures. A mixture of iodosobenzene and iodoxytoluene, or of iodosotoluene and iodoxybenzene, when shaken with silver hydroxide, gives a yield of the base, isolated in the form of i t s iodide, corresponding to about 85 per cent. of the theoretical quantity. Iodoxybenzene, shaken with silver oxide and water, does not give any base, even after the lapse of three weeks. Igdosotoluene shaken with water and silver oxide gives only very small quantities of ditolyl- iodoniam hydroxide, Hence, to get 80 per cent.of the theoretical quantity, both compounds must take part in the reaction. The principal product obtained from the mixture of iodoso- and iodoxy-compounds is therefore a definite compound, and not merely a mixture of diphenyl- and ditolyl-iodonium salts. The bromocamphorsulphonate prepared from pure phenyltolyl- iodonium iodide does not give ditolyliodonium iodide when it is decom- posed with potassium iodide. Therefore, the mixed base can be con- verted into the bromocamphorsulphonate, and this salt can be crystal- lised without change. For the purpose of this investigation, i t was necessary to prepare the bromocamphorsulphonate of diphenyl- and ditolyl-iodonium hydr- oxides ; these salts are described later, and i t may be noted here that whereas diphenyl- and phenyltolyl-iodoniu m bromocamphorsulphona t es are isomorphous and very difficult to distinguish from one another, the ditolyl salt is usually obtained in long needles or prisms, absolutely different from the dodecahedra of the other two salts; it is, however, dimorphous, and under certain conditions crystallises in dodecahedra, indistinguishable by inspection from those of the diphenyl salt.E XP E R I MENT A L. Phenyl-p-tolyliodonium Iodide, z$i>I*I. 0 4 Molecular proportions of iodosobenzene and iodoxy t ol uene were shaken with one molecular proportion of silver oxide and 200 C.C. of water for 36 hours; a t the end of this time, all the yellow powder had disappeared, and the liquid had a strong aromatic odour and a very pale yellow colour.It had, however, very slight basic properties, as practically the whole of the base which is produced interacts with the silver iodate, giving phenyl-p-tolyliodonium iodate. The solution was now filtered from the silver oxide diluted to about 400 C.C. with water, and sulphur dioxide passed in until the precipi- tated iodide had become perfectly white. The whole was then warmed on the water-bath until quite free from1354 PETERS: IODONIUM COMPOUNDS OF THE TYPE IR’R’‘R”’ sulphurous acid, the precipitate separated by filtration, and well washed. To the filtrate, a very dilute solution of potassium iodide was then added, in order to convert any sulphate of the base into iodide. The mother liquors were finally evaporated nearly to dryness, but only a very small quantity of the iodide mas deposited, thus showing that it is very sparingly soluble in cold water.The total yield of iodide was 80 per cent, of the theoretical, theloss being ’principally due to the fact that in the reduction of the iodate some of the base splits up into benzene, toluene, and the iodide of the base.” The iodide was purified by recrystallisation from dilute alcohol, from which i t separated in lustrous needles melting and decomposing at 153-154’. I t s melting point, however, depends to some extent on the rate of heating, and may be as low a s 152’ or as high as 1 5 8 O , according as the temperature is raised very slowly or very quickly. The iodide is very sparingly soluble in water, chloroform, acetic acid, or absolute alcohol, but dissolves more freely in hot aqueous alco h 01.An iodine estimation was made with the following result : 0.2750 gave 0.3064 AgT. I= 60.16. C13H,,I, requires I = 60.1 8 per cent. A p-tolplphenyliodonium iodide was now prepared by treating iodosotoluene and iodoxybenzene with moist silver oxide, under exactly the same conditions as those described in the previous experiment. It was isolated in the form of its iodide, and recrystallised from dilute alcohol, from which i t was deposited in needle-shaped prisms. An iodine determination gave the following result : 0.3094 gave 0.3159 AgT. C,,H,,T, requires I = 60.18 per cent. When examined under the microscope, this iodide appeared to be identical with the first preparation and its general behaviour towards solvents was the same as that of the previous one.Its melting point was 153’ to 154’ (slightly variable with rate of heating) and when mixed with the iodide previously prepared the melting point remained unchanged. From these facts, it was concluded’ that the iodides prepared by the two differen: methods are identical. I = 60.21.AND THE CONFIGURATION OF THE IODINE ATOM. 1355 Pherzyl-p-tolyliodoi~m Nitrate, #~~>I*NO,. 6 4 To prepare the nitrate, the iodide was treated with one molecular proportion of silver nitrate in aqueous alcoholic solution, when a pre- cipitate of silver iodide was immediately thrown down, The whole was boiled on the water-bath for fifteen minutes, filtered, and the solution evaporated to a small bulk. The nitrate, which quickly solidified, was then recrystallised from dilute alcohol.It was thus obtained in short needles, melting at 117", which were very soluble in aqueous alcohol. An iodine estimation was made with the following result : 0.3761 gave 0,2349 AgI. I= 34.92. CI,H,,I*NO, requires I = 34.04 per cent. Phenyl-p-tolyliodoium Bromocamphorszcl~honate, C,H,Me C6H5>I*S0, C,oH,,OBr. The iodide of the base which had been previously prepared from iodosobenzene and iodoxytoluene was treated in alcoholic solution with silver bromocamphorsulphonate in molecular proportion ; an immediate precipitation of silver iodide occurred and the whole was warmed gently on the water-bath for some time in order to complete the reaction. The solution was then filtered and left to evaporate at the ordinary temperature.After about four days, it began to deposit crystaIs and crystallisation was continued until about 2 grams of the salt had separated out; the mother liquor was again left to crystallise and so on until several successive crops of crystals had been obtained. These crystals of phenyl-p-tolyliodonium bromocamphorsulphonate consist of well-defined, highly lustrous, dodecahedra which may be grown to a considerable size. They contain water of crystallisation, and in consequence have not a definite melting point, but begin to soften at about 105' and liquefy completely at about 120'. A determination of the water of crystallisation gave the following result : 0.6062 lost 0.0220 H,O at 100". C2,H,604SBrI,H,0 requires H,O = 3.6 per cent. This result agrees with that required for 1 mol.of water, but after having been heated a t 100" for some hours, the salt is somewhat sticky and faintly brown, H,O = 3.6.1356 PETERS: IODONlUM COMPOUNDS OF THE TYPE 1R'H''R'" A halogen determination with the anhydrous salt was made with 0.21 12 gave 0-1476 AgBr + AgI. C,,H,60,SBrI requires Br + I = 34.2 per cent. The anhydrous salt has not a very definite melting point, and softens a t about 162', melting completely a t about 165' when heated quickly. Phenyl-p-tolyliodonium bromocamphorsulphonate crystallises well from dilute acetone, the crystalline deposit being identical with that obtained from dilute alcohol. It is very readily soluble in alcohol or acetone, but only sparingly s3 in ethyl acetate or chloroform, and very sparingly so in cold water.When warmed, it has a very peculiar, rather pungent, highly characteristic odour. XpeciJc Rotation.-Owing to the well-defined character of the crystals of this salt, there was little hesitation in arriving a t the conclusion that the various deposits, which were obtained without difficulty in the manner described above, were identical, and that fractional crys- tallisation of the salt had failed to resolve i t into the salts of two different bases ; this conclusion was confirmed by the following deter- minations of the specific rotations of the various fractions of the salt ; I. 0.5 gram of air-dried salt was dissolved in aqueous methyl alcohol, the solution diluted to 25 C.C. with water, and examined in a 200 mm. tube ; the aean of several concordant readings gave a + 1°48', hence 11.0.6062 gram, under the same conditions, gave a + 2'12', hence IIIa, 0.5 gram of the air-dried salt was dissolved in pure methyl alcohol, and examined under the same conditions as previously ; the mean of several concordant results gave a + 1'58', hence [ a ] D + 49.1'. 11171. 0.5 gram dissolved in water and very little methyl alcohol gave These experiments gave practically the same values for the specific rotation in the case of all the fractions, and taking the mean value of the specific rotation of the air-dried salt (I, 11, and IIIb) as [a], + 45.6', that of the anhydrous compound would be [a] + 46.7' ; from this value the molecular rotation of the salt may be calculated to be [ M I D + 282'. Now the molecular rotation of bromocamphorsulphonic acid is [&I], +270°, hence the above result seems to prove that the base of the salt is practically inactive, assuming that in aqueous methyl alcoholic solution the salt is dissociated to a sufficient extent to give the true molecular rotation of the acid ion.That this is probablythe case may be inferred from the results of the experiments with fraction IIIa! which show that in aqueous methyl alcoholic solution containing the following result : Br + 1 = 34.2. [ U ] D + 45'. [a]D + 45O. a f 1°52', hence [ a ] D + 46.6'.AND THE CONFIGURATION OF THE IODINE ATOM. 1357 very little methyl alcohol the specific rotation is only a little lower than in the case of the solution in anhydrous methyl alcohol. As, however, this might not be the case and as the specific rotation in aqueous solution could not be determined owing to the slight solu- bility of the salt, it seemed desirable to continue the fractional crystallisation until as large a proportion of it as possible had been obtained in a state suitable for examination.On doing so, it was observed that from the last mother liquors a con- siderable proportion of the salt was deposited as a gum from which we were unable to isolate a pure preparation, These observationsseemed to shorn that the original specimen which was thought to be pure phenyl-p-tolyliodonium bromocamphorsulphonate had been resolved into two different fractions; on the other hand, the results of the optical examination of the first deposits indicated that the most sparingly soluble salt was not that of an optically active base.I n order to try and ascertain the nature of the salt in the last mother liquor, small portions of the most sparingly soluble and of the most readily soluble fractions were separately decomposed with potass- ium iodide, whereupon the iodide of the base was precipitated in each case. The iodides appeared to be identical in most of their properties, and they both crystallised in fine, silky needles from dilute alcohol, but the first fraction melted a t about 170' to 175' when heated very quickly, whilst the other melted very indefinitely at about 154'. Very little reliance, however, could be placed on these melting points, as they were not at all sharp in either case, and varied very consider- ably with the rate of heating. Neither iodide gave any divergence to polarised light, but owing to their very slight solubility only extremely dilute solutions could be used.In order to try and determine the nature of the impurity in the phenyl-p-tolyliodonium bromocamphorsulphonate, the p-tolylphenyl- iodonium bromocamphorsulphonate, prepared from iodosotoluene and iodoxybenzene, as described above, was fractionally crystallised. The highest fraction again appeared homogeneous, and identical with the phenyl-p-tolyliodonium bromocamphorsulphonate in physical, optical, and chemical properties, The last mother liquors gave a sticky, gummy mass which crystal- lised with some difficulty from dry ethyl acetate in long needles or prisms melting at about 185O. Portions of the first and last fractions of this salt were separately decomposed with potassium iodide.The iodides thus formed, recrys- talliaed from alcohol, were not identical ; the one from the least soluble portion crystallised in silky needles, whilst that from the most soluble1358 PETERS: IODONIUM COMPOUNDS OF THE TYPE 1R’R”R”’ portion crystallised in octahedra. They melted a t 154’ and 143’ re- spectively, but the melting points were very variable, differing by 10’ or 15’according to the rate of heating. The iodide from the first fraction was also much less readily soluble in alcohol, ethyl acetate, or water than that from the most readily soluble fraction. The discovery of these two salts proved, therefore, that the original phenyl-p-tolyliodonium bromocamphorsulphonate was a mixture.Now, since iodoso-derivatives very readily undergo decomposition when boiled with water or alkali, it seemed probable that such a decomposition might have occurred in the preparation of the mixed base ; using a mixture of iodosotoluene and iodoxybenzene, for example, the former might give iodoxytoluene, which would then interact with the iodosotoluene giving ditolyliodonium hydroxide. In order, therefore, t o test this conclusion, p-ditolyliodoniurn iodide was prepared and its properties studied. Di-p-tolyliodonizcm Iodide, (C,H,Me),~*T. This compound was prepared by shaking iodoso- and iodoxy-toluene in molecular proportion with silver oxide and water. The iodide was then isolated in the usual manner and orpstallised from dilute alcohol, in which it was much more readily soluble than diphenyliodonium iodide.It was thus obtained in octahedral crystals of a very lustrous appearance, melting at 143’ to 156O, according to the rate of heating. An estimation of the halogen was made with the following result : 0*2105 gave 0.2280 AgI. I=58.5. CI4Hl4I2 requires I = 58.3 per cent. In appearance, crystalline shape, melting point, and general be- haviour, this iodide was identical with that obtained by precipitating the most soluble portion of the p-tolylphenyliodonium bromocamphor- sulphonate, and yet so entirely unlike the main portion of that salt that we bad no hesitation in concluding that the latter contained the ditolyl salt as impurity. Di-p-tolyliodonium Brornocamnp~orsul~~onate, (C6H4Me)21’S03’C,oH,*OBr. This salt was prepared from the iodide and silver bromocamphor- It was thus sulphonate and crystallised from alcoholic ethyl acetate.obtained in long needles melting a t 185-186’. A halogen estimation was made with the following result : 0-4034 gave 0.2788 AgBr + AgI. Br + I = 33.8. C,,H,,O,SBBrI requires Br + I = 33.4 per cent,AND TEE CONFTGURATION OF TEE IODINE ATOM. 1359 This salt appeared to be dimorphous, and when crystallised from aqueous alcohol it sometimes separated in dodbcahedra which resembled those of phenyl-p-tolyliodonium bromocamphorsulphonate, and, like the latter, probably contained water of crystallisation. Having learnt the properties of the ditolgliodonium bromocamphor- sulphonate, a fresh investigation of the last mother liquor of the tolyl- phenyliodonium bromocamphorsulphonate led to the isolation of a small quantity of a salt which was proved.to be identical with this ditolyl derivative. I n order to ascertain whether the pheny1.p-tolyliodonium bromo- camphorsulphonate contained the diphenyl base, as seemed very prob- able, the latter was prepared in the usual way and isolated in the form of its iodide which was found to melt a t 168--175O, depending upon the rate of heating (Victor Meyer and Hartmann give 175-176O as the melting point). The iodide was treated, as described above, with one molecule of silver bromocamphorsulphonate and the product recrystallised from alcohol, The salt contained water of crystallisation, as shown by the following determination : 0,8284 lost 0,278 H,O a t looo. A halogen estimation with the anhydrous salt gave the following 0.6051 gave 0.4309 AgBr + Agl.C2,H2,0,SBrI requires Br + I = 35.0 per cent. This snIt crystallised well from dilute alcohol in beautiful, lustrous, well-defined dodecahedra melting at about 165-1 6 8 O when previously dried a t 100'; it also crystallised well from dilute acetone, some of the crystals growing to a good size. It was much less soluble in dilute alcohol than the corresponding ditolyl derivative. On the other hand, it was so similar to phenyl-p-tolyliodonium bromocamphorsulphonate in crystalline form and in all other respects that, except for its rather higher melting point, it would be hard to distinguish it from the latter. For this reason, it was not possible to prove beyond all doubt that the original phenyl-ptolyliodonium bromocamphorsulphonate from iodosobenzene is a mixture of a large percentage of phenyltolyl deriv- ative and a very small percentage of the diphenyl compound, but it seems extremely probable that this is so, H,O = 3.35.C,,H2,0,SBrJ + l-$H,O requires H,O = 3.6 per cent. result : Br + I = 38.4.1360 PETERS : IODONIUM COMPOUNDS OF THE TYPE IR’R’”’’’ Proofs of the Existence of a Mixed Iodoniurn Base. Having proved that one preparation, which was at first thought to be a pure iodonium base, contained diphenyliodonium hydroxide whilst the other contained the corresponding ditolyl base, doubt arose as to the actual existence of a mixed base, and it seemed not impossible that the substance described as the latter might consist of a mere mixture of diphenyl and ditolyl compounds.The following experiments mere made to settle this point. The diphenyl- (melting point about 165-168’) and the phenyltolyl- iodonium bromocamphorsulphonates were directly compared, but with the exception of a slight difference in melting points (3’ or 4’), the two compounds seemed to be identical; a -mixture of the dehydrated salts melted very indefinitely from 153-156’ and softened about 8’ before the phenyltolyl derivative. The iodides of the diphenyl and phenyltolyl bases were examined in a similar manner. Here again the results were not very conclusive, and the melting point of the mixture of the two salts mas only about 5’ lower than that of the phenyltolyl derivative. The dichromates were next examined and were found to have rather more sharply defined melting points tban the other salts.The three compounds were directly compared, with the following results : Salt of diphenyl base crystallised in prisms melting a t 157’ when slowly heated. Salt of ditolyl base crystallised in plates melting a t 140’ when slowly heated. Salt of phenyltolyl base crystallised in prisms melting at 143’ when slowly heated. Action of Silver Oxide on lodoso- and lodoxy-derivatives. As the experiments just described did not settle satisfactorily this question of the existence of a mixed base, the action of silver oxide on an iodoso-compound alone and on an iodoxy-derivative alone mas studied. Iodoxy- and iodoso-benzene were shaken separately with silver oxide and water. After a few hours’ shaking, the iodoso-derivative gave some iodide when treated in the usual way, but the iodoxy-benzene, even after prolonged shaking, did not give tho slightest trace.Now when a mixture of iodoso- and iodoxy-derivatives is used, the yield of the iodide, as already stated, is 80 per cent. of the theoretical. The experiment just described, therefore, proves that the mixed base is formed by the cooperation of both the iodosobenzene and iodoxy- toluene, and consequently it must be a phenyltolyl derivative,AND TH'E: CONFIGURATION OF THE IODINE ATOM. 1362 Examination of Iodides. -We further proved by direct experiment that a mixture of the iodides of the diphenyl- and ditolyl-iodonium bases is easily resolved into its components by fractional crystallisation. This was done by first dissolving equal proportions, then a large proportion of diphenyl and a small proportion of ditolyl iodides in dilute alcohol and fractionally crystallising.I n both cases, the ditolyl derivative came down in the last fractions and was easily detected. When, on the other hand, the iodide of the pure mixed base was frac- tionally crystallised, not the slightest trace of the ditolyl derivative could be found, The identification of the impurities found in the p-tolylphenyl and phenylp-tolyl bromocamphorsulphonates as the ditolyl and the diphenyl derivatives respectively, and the proof that they exist in only small quantities, due to the formation of the iodoxy- from the iodoso-deriv- ative, and not from any intramolecular change, makes it evident that the mixed base is a definite compound and not a mere mixture of diphenyl- and ditolyl-iodonium bromocamphorsulphonates. Phelnyl-p-tolyliodonium Chloride, C6H4Me>I*Cl. C6H5 This salt was obtained by dissolving the bromocamphorsulphonate in a little aqueous alcohol and adding an excess of an aqueous solution of sodium chloride ; on cooling, the sparingly soluble chloride crystal- lises out in colourless needles. It was also prepared by treating the original mother liquors, which contain the base and the iodate of the base, with hydrochloric acid, when chlorine is evolved and the chloride precipitated. It crystallises well from dilute alcohol in needle-shaped crystals melting at about 193'. A halogen estimation was made with the following result : 0.1 186 gave 0.135 1 AgCl + AgI. C13Hl,ClI requires C1+ I == 49.1 per cent. P~enyl-p-tolyliodonium Dichromate, is easily obtained by dissolving the brornocamphorsulphonate in a little warm alcohol and adding an aqueous solution of potassium dichromate until a slight turbidity is formed. On cooling, the salt crystallises in thin prisms of a beautiful, bright red colour. It melts at 143' and decomposes a t a slightly higher temperature. UXIVERSITY COLLEGE, NOTTINGH AM. C1+ I = 48.9.