THE INFLTJENCE OF LIGHT ON DIAZO-REACTIONS. I. 85 1V.--The I n j u e n c e of Light orb Diazo-reactions. I. By KENNEDY JOSEPH PREVITE ORTON, JOSEPH EDWABD COATES (and, in part, FRANCES BURDETT). THE voluminous literature of the diazo-group does not indicate that the remarkable effect of light on certain reactions of this class of compounds has been closely investigated, notwithstanding the fact that more than one '' diazo-type " photographic process has been pat en ted. * * Feer (D.R -P. 53455) patented a process in which a film, coated with a mixture of a diazosulphite and a phenol or an amine, was exposed to light. A decomposition of the former occurred, which was followed by coupling with the phenol or amine, and hence the production of a coloured negative. The unchanged sensitive material was washed away after exposure.Green, Cross, mid Bevitn (I>.R.-P. 56606 ; Ber., 1890, 23, 3131 ; and J. SOC. C'/mn. Ind., 1890, 9, 1001) brought forward a method for the use of the diazo- tlcrivative of priiiiuline for a similar purpose. A " negative '' was obtained by ex- posing films coated with the diazo-compound, the decomposition of which was pro- portional to the intensity of the light. A '. positive " was developed by treatment with iin amine or a phenol. I n regions of faint illumination, where the diazo-com- pouiid had not been deeoinposed, a colour developed ; where the illumination had been intense, all the diazo-compound had been decomposed and the positive was colourless. They further established that the red end of the spectrum was the more active, and that nitrogen was evolved from the diazo-compound.They, however, express themselves as " undecided whether the product is a phenol, . . . . or whether the primuline residue enters, the molecule of cellulose." They conclude, moreover, " that molecular union with the medium is a necessary condition, . . . . for the free dinzopriniuline, when exposed t o light in a thin film is either not decomposed at all, or only after very prolonged exposure.') t Ruff and Stein (Be?., 1901, 34, 1668), using a similar photographic method, have 0 236 ORTON, COATES, AND BURDETT: THE INFLUENCE OF I n the course of an investigation of some reactions of s-trisubsti- tuted diazobenzenes, notably s-tribromodiazobenzene (Orton, Trans., 1903, 83, 796; 1905, 87, 99), it was observed that the diazonium salts, the hydrogen sulphate, and the nitrate, both as solid and in solution, were exceedingly sensitive to light.The instability was the more remarkable, inasmuch as this type of diazonium salt is singularly permanent at the ordinary temperature. The effect of exposure to light could accordingly be most easily demonstrated and studied in the case of such s-trisubstituted diazonium salts. The nature of the decomposition of the diazonium salt depends on the solvent. I n water, a phenol is formed; in methyl alcohol, a methyl ether, ArnObCH, j in ethyl alcohol, an ethyl ether, Ar*O*C,H,, and in acetic acid, the phenyl acetate, Ar*O*CO*CH,; thus, for example : Ar*N(HSO,)tN + CH,'CO,H = Ar*O*CO*CH, + H,SO, + N,, If the salt is a chloride or a bromide, the diazo-group is also replaced to some extent in aqueous solution by chlorine or bromine.The influence of light is well illustrated by the behaviour of dilute solutions of 5-bromo-m-xylene-, or 6-bromo-~-cumene-diazonium hydrogen sulphates. Solutions of these salts can apparently be pre- served indefinitely at the ordinary temperature if light is rigidly excluded; thus in ten weeks no measurabIe volume of nitrogen was evolved from a 1 per cent. solution of either of these salts, If such solutions are boiled, transformation to the corresponding phenols is rapid and quantitative. Exposure to diffused daylight is followed by evolution of nitrogen; in direct sunlight, the speed of the reaction is considerable and the yield of phenol quantitative, The case of s-tribromodinzobenzene is of particular interest, inasmuch as it has been shown, by all who have investigated this substance, to behave abnormally when its solutions in water, in methyl or ethyl alcohol, or in acetic acid are heated. Silberstein (J.pr. Chem., 1883, 27, 113) could isolate no s-tribromophenol in the decomposition of aqueous solution of the diazonium nitrate. Hantzsch (Bey., 1900, 33, 2517) confirmed this observation, and, in addition, ascertained that no s-tribromophenyl methyl- or ethyl-ether was formed on heating the solutions in the corresponding alcohols, s-tribromobenzene being the sole product. Similarly, he detected no s-tribromophenyl acetate in the reaction with acetic acid, s-tribromobenzene again being alone isolated.The changes which occur when aqueous solutions of these tribromo- investigated the effect of the constitution of the diazo-compound on its " sensitive- ness." O n the basis of some experiments of Andresen (Photogrcqhische Corre- spo7~dei~z, 1595), they conclude that the diazo-compound midergoes the phenolic decomposition.LIGHT ON DIAZO-REACTIONS. I. 37 benzenediazonium salts * are heated were first described by one of us (Trans., 1903, 83, 802); i t was found that the diazonium salt mainly decomposed into dibromoquinonediazide, bromine appearing in the ionic condition, thus : C,H,Br,*N(HSO,)iN + H,O = O,:C,H,Br,:N, + HBr + H,SO,. I n the course of this research, this reaction has been again investi- gated, the diazonium salts being now also heated in solution in 63 per cent.sulphuric acid, a method first employed by Heinichen in the preparation of 2 : 6-dibromophenol (Annulen, 1889, 253, 281). Although such s-trisubstituted benzenediazonium salts, as those obtained from 3 : 5-dibromo-o-toluidine and 3 : 5-dibromo-p-toluidine, which do not undergo the phenolic decomposition when their aqueotis solutions are heated, are nearly quantitatively converted into the corresponding cresols, if 63 per cent. sulphuric acid is used as a medium, yet s-tribromodiazobenzene is still refractory. The diazonium hydrogen sulphate was rapidly decomposed, but no s-tribromophenol was formed. Subsequent to the publication of the preliminary notice (Proc., 1905, 21, 168), Gain and Norman (Proc., 1905, 21, 206) showed, however, that some 2 per cent.of s-tribromophenol could be obtained if the method described in the German patent (D.R.-P. 95339), namely, heating the diazonium salt with dilute sulphuric acid and sodium sulphate, were used. This result has been confirmed by the authors, who have isolated small quantities of s-tribromophenol from among the products of decomposition of s-tribromobenzene- diazonium hydrogen sulphate, when it is treated according to the directions in the patent. The main product of the reaction is a material resulting from a transformation of the quinonediazide, which is itself not permanent under the conditions. The decomposition, in fact, under these conditions mainly follows the course of that of the aqueous solution. I n the face of such observations it was scarcely to be expected that a means of converting s- tribromodiazobenzene quantitatively into s-tribromophenol could be found. Nevertheless, such a complete con- version of the diazonium salts in aqueous solution is induced by light.Further, both the s-tribromophenyl methyl- and ethyl-ethers and s-tribromophenyl acetate are produced if solutions in methyl or ethyl (90 per cent.) alcohols or acetic acid are exposed to light. The yield of the phenyl acetate is quantitative, but under the most favourable conditions some 30 per cent, of s-tribromobenzene is formed together with the ethers. Solutions of the diazonium salt in methyl o r ethyl * The hydrogen snlphate mas mainly used, since there is sonie difficulty in prc- serving the nitrate ; further, the chloride cannot be easily isolated, an acid chloride being precipitated from the acetic acid solution by ether.38 ORTON, COATES, AND BURDETT: THE INFLUENCE OF alcohol differ in some respects.Whilst the solution in methyl alcohol remains unchanged in the dark, that in ethyl alcohol slowly decom- poses, but, in marked contrast to the decomposition induced by light, now only s-tribromobenzene is formed. Solutions (or suspensions) in several other media have been examined. When dissolved in 95 per cent. sulphuric acid, the phenolic decomposition takes place slowly, no other substance being produced. Solutions in fuming nitric acid appear to behave similarly, but here a secondary reaction, namely, displacement of the bromine by the nitro-group, complicates the phenomenon.Formic acid solutions yield only s-tribromobenzene, thus : C,H2Br3*N(HS0,)IN -t H*CO,H = C,H3Br3 + H,SO, + N, + CO,. I n propionic acid, the diazonium salt is insoluble ; nitrogen is evolved from the suspension, but the reaction is exceedingly slow. The diazonium salts of 2 : 4 : 5 : 6-tetrabromodiazobenzene, 2 : 6-di- bromodiazobenzene, 2 : 4 : 6-tribromo-3-nitrodiazobenzene, 3 : M i - bromo-p-diazotoluene? and 3 : 5-dibromo-o-diazotoluene, in so far as they have been studied, behave in a completely analogous manner. Aqueous solutions of benzen ediazonium salts ra.pi d 1 y clian ge when exposed t o sunlight a t 0" ; nitrogen is evolved and phenol formed, but the decomposition soon comes to ;L standstill, owing to a dark turbidity arising frcm some secondary reaction which prevents access of light.This diaculty is not met with in the case of $-cumenediazonium salts, which are, moreover, even less stable in aqueous solutions than benzenediazonium salts a t the ordinary temperature. When exposed t o light a t O", nitrogen is rapidly evolved, +-cumenol being formed, whilst if shielded from light this decomposition is very slow a t that Iow temperature. In marked contrast t o the sensitiveness of the diazonium salts is the stability of solutions of diazotates. Aqueous solutions of sodium p-nitrobenzenediamtate, 3 : 5-dibromo-p-toluenetZiazotate, and benzene- isodiazotate are unchanged after two or three days' exposure to light. A solution of potassium s-tribromobenzenediazotate in methyl alcohol is equally stable. Investigation of the effect of -variation in the concentration and nature of the acid on the decompositions of diazonium salts, which are accelerated by light, has shown that the former (concentration of the acid) has remarkably little influence on the rate of the transformation.Even up t o concentrations of 25 per cent., sulphuric acid does not diminish the speed; a t a concentration of 30 per cent., a slight decrease is perceptible, but, as mentioned in the foregoing, even solutions in 95 per cent, sulphuric acid yield phenol on exposure to light. The nature of the acid is only of consequence when the acidLIGHT ON DIAZO-REACTIONS. I. 39 radicle can itself replace the diazo-group ; in the presence of hydrogen chloride or bromide, light hastens not only the conversion into phenol, but also the replacement of the diazo-group by chlorine or bromine. The most marked effect of the presence of acids is seEn in the case of those halogendia zobenzenes which tend to undergo another decom- position, namely, the formation of a quinonediazide with the elimination of halogen.This reaction is markedly accelerated by light, and con- sequently takes place in solutions of diazonium salts exposed to light concurrently with the phenolic decomposition ; the latter, however, is always the dominant reaction. Thus i t was found that from a 1 per cent. solution of s-tribromobenzenediazonium acetate, the amount of bromine eliminated in a given time on exposure to light was three times that set free in a similar solution kept in the dark. No phenol was formed in the latter solution, whereas, in the former, 100 molecules of diazo-compound yielded phenol for every 77 which lost an atom of bromine.This acceleration of the quinonediazide reaction is still perceptible in solutions of the diazonium hydrogen sulphate. I n a 1 per cent. solution of the diazonium hydrogen sulphate, the pro- portion of molecules yielding phenol to those giving quinonediazide is 100 : 25, in a 15 per cent. solution of sulphuric acid i t has fallen to 100 : 12, whilst i n a 30 per cent. solution the elimination of bromine is no longer perceptible. The rate a t which the decomposition proceeds on exposure to light is greatly influenced by the extent to which the quinonediazide reaction is occurring, for the reason that, when exposed to light, all quinonediazides rapidly change into deeply-coloured, amorphous solids, which, remaining suspended in the solution, effectually prevent the access of light.According to the view previously explained (Orton, Trans., 1903, 83,796), the quinonediazide is the result of an interaction of diazonium {Ar*NiN)” and hydroxyl (OH)‘ ions. As the concentration of acid increases, the concentration of the hydroxyl ions becomes vanishingly small, and the formation of quinonediazide ceases. This change accompanying the replacement of the diazo-group by hydroxyl was observed to a greater or less degree in the case of all the halogendiazobenzenes. It is s-carcely detectable, if not entirely absent, in the decomposition of 5-bromodiazo-nz-xylene and 6-brcmodi- azo-$-cumene, is noticeable in that of the two 3 : 5-dibromodiazo- toluenes, and assumes still larger proportions as the number of negative bromo- and nitro-groups in the benzene nucleus is increased, until in the case of an aqueous solution of 2 : 4 : 6-tribromo-3-nitro- benzenediazonium hydrogen sulpha.te 46- 47 per cent.of one atomic proportion of bromine is eliminated.40 ORTON, COATES, AND BURDETT: THE INFLUENCE OF A suggestion as to the mechanism of the transformation of diazonium salts into phenols, which has attained considerable acceptance, was first given by Hantzsch (compare ‘‘ Diazo-Verbindungen,” Ahrens’ Sammlung), who represents the phenols as being primarily formed from the syn-diazohydroxides, thus : Ar-8 Ar HO--N -+ 4~ -k Nv by an ‘‘ intramolecular decomposition ” which is comparable to the decomposition of t,he syn-aldoximes into water and nitriles.Since solutions of the diazonium salts are particularly liable to this reaction, he accounts for the syn-diazohydroxide, which must on this view be present in such solutions, by the existence of a certain amount of hydrolysis of the diazonium salt. The free diazonium hydroxide formed in this manner has been shown (Hantzsch and Davidson, Ber., 1898, 31, 1612) to change partially into the isomeric sp,-diazohydr- oxide, so that an equilibrium exists which can be represented thus : The hydrated compound is suggested as an intermediate step. It should follow, therefore, that all conditions which favour hydro- lysis, for example, diazonium salts of weak acids or the salts of a weak diazonium base, should increase the speed of the phenolic decom- position; further, the presence of an excess of a strong acid should prevent or a t least decrease the rate of this transformation.It was shown, however, by Euler (Anncden, 1902, 325, 292) that the rate of the conversion of diazonium salts into phenols was inde- pendent of the presence of excess of acid, a t least in dilute aqueous solution. Euler maintains, therefore, that the conversion into phenol can be a purely diazonium reaction, and is not confined to sp-diazo- hydroxides. Recently (Bey., 1904, 37, 1087), Hantzsch has accepted this correction, and states that he is now of Euler’s opinion. The experiments described in this paper afford independent evidence for the view that diazonium salts in aqueous solution are directly transformed into phenols.As has been shown, the presence of a considerable excess of acid is no hindrance to the change ; under the influence of light the reaction takes place at the ordinary temperature in 30 per cent. sulphuric acid, with no decrease of speed. Moreover, it will even proceed, though but slowly, in a 95 per cent. solution of sulphuric acid; in such a medium, it is difficult to imagine even a vanishing trace of hydrolysis. Further, in the light of the experiments here recorded, there would appear to be little reason to think that the other decompositions, namely, the reaction with methyl or ethylLIGHT ON DIAZO-REACTIONS. I. 41 alcohols or with acetic acid, which lead respectively to ethers and acetic esters, are reactions of syz-diazo-compounds and not of diazonium compounds.These transformations appear to be regarded by Hantzsch (Zoc. cit., p. 6s e t seq.) as decompositions of a syn-diazo-compound. The presence of excess of acid is, in these reactions, however, not without effect, in that the product of the reaction induced by light is modified ; whereas, under all conditions yet investigated, s-tribromobenzene always accompanies the s-tribromophenyl methyl- and ethyl-ethers, the proportion of the former is increased in strongly acid solution. I n any case, the fact that excess of acid favours the replacement of the diazo-group by hydrogen affords strong evidence that this transforma- tion, a t least, is a reaction of the diazonium compound.Possibly t h g experiments which are now in progress on the accelerating influence of light on the replacement of the diazo-group by chlorine or bromine may illuininate further the mechanism of these diazo-reactions. With regard to the mechanism of the action of light, it may be suggested that the solvent becomes associated with the diazonium salt (or in dilute solutions, with the diazonium ion). Considering how freely formulz for hypothetical intermediate products in diazo-reactions have been brought forward in recent years, it is, perhaps, hazardous to attempt to represent graphically this additive product, But in any case it would seem that the quinquevalent " basic " nitrogen at,om, which is linked t o the acid radicle, cannot be directly involved, or undergo a change of valency, if the reactive compound is a diazonium , 1s derivative. Ar*Y*(HSO,) .OH*N*H From this point of view the expression, Ar*N( HSO,) not admissible, whereas the formula, , in which the fi ( H20) residual valency of the tervalent nitrogen is brought into play, is open to less objection. Such a complex may be supposed to be unstable, breaking up either under the influence of light or heat, yielding a phenol and nitrogen,% Further, it may be supposed that in the case of certain diazonium salts, such as those with several negative. substituents, s-tribromodiazobenzene and the like, this unstable com- plex may be resolved on mere heating into its constituents, water and diazonium salt, a fact which would account for the difficulty of con- verting the diazo-compound, just named, into the corresponding phenol.From the foregoing it is obvious that if such a complex exists, its formation is not materially affected by the presence of acid, The reactions of s-tribromobenzenediazonium hydrogen sulphate AP'N *( HS04) * Perhaps in solution in snlphuric acid, a complex is the sensitive #( H,S04) - - material.42 ORTOK, COATES, AND BURDETT: THE INFLUENCE OF l N Z N + 1 and &fN + I .LIGHT ON DIAZO-REACTIONS. I. 43 Solutions in methyl or ethyl (90 per cent.) alcohols or in acetic acid are equally sensitive, although, owing t o the solubility of the products of decomposition, the rapid change is only made obvious by the evolution of nitrogen when the flask containing the solution is con- nected with a nitrometer. In the case of this and similarly constituted diazonium salts, the decomposition of the aqueous solution pursues two courses, the main reaction results in the formation of s-tribromophenol, nitrogen being evolved ; a t the same time, however, bromine is eliminated in the ionic condition, and a quiiionediazide produced. The last-mentioned com- Found is itself changed by light, being transformed into a dark brown, amorphous powder.For the reasons given in the introduction, the presence of excess of acid should reduce the proportion of the subsidiary reaction whatever effect such an increase in acidity had upon the phenolic decomposition. The latter change can therefore be best followed in acid solution. On exposing a 1 per cent. solution of s-tribromobenzenediazouium hydrogen sulphate in 15 per cent.sulphuric acid to direct sunlight, a white solid which consisted of minute crystals was deposited, whereas in the absence of the acid a dark brown material separ- ated. The solid was collected, washed, and extracted with sodium carbonate, which left, in the case of the aqueous solution, a considerable quantity of brown solid undissolved. The phenol was precipitated from its solution in alkali and dissolved in dilute alcohol, from which it crystallised in long, slender needles melting a t 92' ; its melting point was unchanged by admixture with a specimen of s-tribromophenol : 0.1925 gave 0.3262 AgBr. CGH,ORr, requires Br = 72.50 per cent. Efect of the Concentration of the Xulphuric Acid o n tihe Decomposition of the Diaxoniuna Salt unde?* the Injluence of Light.-The salt appears to decompose most rapidly, as estimated by the evolution of nitrogen, in the presence of a 15-30 per cent, sulphuric acid, A t lower con- centrations, the evolution of nitrogen is slower, but the retarding influence may be due to the dark precipitate, which acts simply by preventing the light from gaining access to the liquid.At higher concentrations, the decomposition is again retarded, but the phenol is formed even in solutions in concentrated (95 per cent.) sulphuric acid. The following results will show the quantitative relation existing between the two reactions, the elimination of bromine and the phenolic decomposition, in different concentrations of sulphuric Br = 72.12. . - acid. (i) On exposing 50 C.C.of a 1 per cent. solution of the diazonium44 ORTON, COATES, AND BURDETT: THE INFLUENCE OF hydrogen sulphate for forty-eight hours (in October), 55 per cent. of the salt was decomposed, 11.8 C.C. of nitrogen were evolved, and the liquid yielded 0.0249 gram of silver bromide. Ratio of the number of molecules yielding phenol and nitrogen to the number giving quinone- diazide and bromine = 100 : 25. (ii) Two exactly similar experiments were made, using a 15 per cent. solution of sulphuric acid instead of water as solvent. ( a ) Eighty-six per cent. of the salt was decomposed, 21 C.C. of nitrogen were evolved, and the liquid yielded 0.0212 gram of silver bromide. ( b ) Sixty-six per cent. of the salt was decomposed, 16.5 C.C. of nitrogen mere evolved, and the liquid yielded 0.0176 gram of silver bromide.The ratio of the two decompositions was in each case 100 : 12. I n these experiments, the volume of the gas could only be approxi- mately measured, and, further, i t is assumed that nitrogen is formed only from the phenolic decomposition, an assumption which was shown to be justified by direct estimation (weighing) of the phenol. The un- changed diazo-compound was weighed as P-naphthol derivative, and hence the percentage of decomposed salt was obtained. Efect of Temperature on the Velocity of Becomposition.-There are great difficulties in the way of determining the velocity of a reaction which depends on the varying illumination of the sun. The difficulties are increased in the case of the phenolic decomposition of the diazonium salts by the precipitation of tribromophenol, which acts as a screen to the solute yet undecomposed. Equal amounts of two 1 per cent.solutions of the diazonium salt in 15 per cent. sulphuric acid were exposed in similar flasks attached to nitro- meters. One flask was immersed in ice and water, and the other in water at a given temperature, great care being taken that each flask was equally exposed to light. In these circumstances all irregu- larities in the illumination affected each solution equally. The progress of the decomposition was estimated by measuring the nitrogen evolved. Several experiments were made ; the following will illustrate the difference of temperature. Experiment 1.-Flask A, temperature 17", 50 per cent. of the diazonium salt was decomposed in two hours fifty minutes; flask B, temperature 2 O , 50 per cent, decomposed in four hours.Experiment 11.-Flask A , temperature 20°, 50 per cent. of the diazonium salt was decomposed in one hour forty-two minutes; flask B, temperature 2O, 50 per cent. of the diazonium salt was decom- posed in two hours twenty minutes. I n this experiment the illumination was more intense than in Experiment I. Action of Light on s-T&%omobenxenediuzotccte.-A solution of the diazotate was prepared by adding 0-25 gram of diazonium salt Attempts were made to ascertain the effect of temperature,LIGHT ON DIAZO-REACTIONS. 1. 45 dissolved in 25 C.C. of water t o 13 C.C. of a 10 per cent. solution of sodium hydroxide. The solution was exposed to sunlight on two successive days.Nitrogen was not evolved nor was bromine eliminated ; the solution, moreover, remained colourleas. Attempts to obtain s- Tribromophenol by Boiling Aqueous Solutions OJ the Diccxonium Sa Its. It has frequently been observed that substituted benzenediazonium salts do not yield phenols, or at least only in small quantities, when the aqueous solutions of their salts are boiled. I n the case of s-tribromobenzenediazonium hydrogen sulphate it has previously been shown that bromine is eliminated, and a dibromoquinonediazide formed, at least in the first instance (Orton, Trans., 1903, 83, 802) ; no s-tribromophenol was discovered. The decomposition of aqueous solutions of this salt has now been more exactly followed. Solutions varying in concentration from 1 to 10 per cent.were heated a t looo’ in an apparatus from which the air had been expelled by carbon! dioxide, so that the nitrogen evolved could be measured. Decomposi- tion was complete in two hours; from 75 to 80 per cent. of the total nitrogen was evolved, the lower number being obtained from the more concentrated solutions. No diazonium salt remained un- decomposed, but the yellow liquid contained quinonediazide, which could be extracted with chlorqform and coupled in the usual way with P-naphthol (compare Orton, Trans., 1905, 87, 104). From 58--60 per cent. of one atomic proportion of bromine was found in the solu- tion. The solid product of the reaction was a brown, amorphous powder, which contained small amounts of dibromoquinonediazide, and s-tribromobenzene.Both these could be extracted with alcohol, but the main part of the solid was insoluble in that solvent. The addition of reduced copper accelerated the evolution of nitrogen, but did not otherwise appear to affect the course of the de- composition. Since it has been noted by Heiniclien (Annalen, 1889, 253, 281) and others that the phenolic decomposition takes place more readily in fairly concentrated solutions of sulphuric acid, for example, in a solution of sulphuric acid boiling at 150°, attempts mere made to obtain s-tribromoyhenol by using 50-63 per cent. sulphuric acid as solvent. No s-tribromophenol could be obtained from the product of decomposition ; but a very little s-tribromobenzene was isolated. In the case of the 50 per cent. acid, only about 3 per cent.of one atomic proportion of bromine was eliminated, but from the 63 per cent. acid, which boils at 150°, as much as 35 per cent. was found, a difference which was probably to be attributed to the higher temperature. Tn46 ORTON, COATES, AND BURDETT: THE INFLUENCE OF all cases, the acid mother liquor was bright yellow, and contained quinonediazide. * Becomposition of the Dry Diazonium ,!?&.-The diazonium hydrogen sulphate (1 gram) was placed in a desiccator containing phosphorus pentoxide, and kept evacuated in the dark for twelve hours. On ex- posure to sunlight, the salt became discolonred, and, after a few hours, of a chocolate-coloured hue. On examination, it was found that the surface only had been affected, the dark solid was insoluble in water, and thus could be freed from the unchanged diazonium salt, It was found to contain halogen but no nitrogen; it was insoluble in alkali hydroxide, and in the usual organic solvents with the exception of glacial acetic acid, in which it could, at least partly, be dissolved.The quantity was to3 small to admit of further investigation, Acceleration of the Elimination of Bromine by Light.--The ex- periments were made with s-tri bromobenzenediazonium acetate, which undergoes the quinonediazide decomposition at a convenient rate (Orton, Zoc. cit.). A 1 per cent. solution of the diazonium hydrogen sulphate (0.25 gram) containing three equivalents of sodium acetate was exposed for five hours to the light i n a flask attached to a nitrometer. A bulky, brown solid separated, and a small amount (4.6 c.c.) of nitrogen was evolved.27.6 per cent. of one atomic proportion of bromine (AgBr= 0.0296 gram) was eliminated. The ratio of the number of molecules undergoing the phenolic decomposition to those undergoing the quinonediazide decomposition is therefore 100 : 77. An exactly similar solution kept in the dark for the same period evolved no nitrogen; the solution became yellow in colour, and a slight yellow precipitate formed. Ten per cent. of one atomic proportion of bromine ( AgBr = 0.0108 gram) was eliminated. Repetitions of the experiment gave a similar result. As was to be expected, variations in the intensity of light affected the phenolic decomposition to a greater extent thaii the quinonediazide decomposition.Decomposition of s-~ribromobenxe.lzedin,yonium II'ydrogen Sulphate in Itf ethyl Alcoholic Xolutim-The diazonium salt is readily soluble in methyl alcohol, and the solution when kept in the dark is stable ; * Dr. J. C. Gain drew the author's attention to the fact that the method of inducing the phenolic decomposition by heating the diazonium salt in a saturated solution of sodiiim sulphate was effective in converting a small proportion of s-tribromobenzenediazonium hydrogen sulphate into s-tribromophenol. An experi- ment was tried, usiug 10 grams of the diazonium salt in a solution made up of 14 C.C. of sulphuric acid and 32 C.C. of water, 18 grams of hydrated sodium sulphate being added. The mixture was distilled in a slow current of steam, the flask being heated in an oil-bath.A small quantity of s-tribromophenol (m. p. 88") mixed with another substance melting a t 75-76', which does not dissolve i n aqueous sodium carbonate, was oltained.LIGHT ON DIAZO-REACTIONS. I. 47 thus no nitrogen was given off from a 1 per cent. solution which was kept in the dark f o r twenty-eight hours (temperature 10.5'); but on exposure to sunlight for one hour 72 per cent. of the salt was decomposed. On again placing this solution in the dark, de- composition ceased. Although solutions in methyl alcohol at the ordinary temperature are stable if screened from light, decomposition rapidly occurs if the solution is boiled, s-tribromobenzene being alone produced. The main decomposition under the influence of light is accompanied by a small amount of-a secmdary reaction, namely, the elimination of bromine, and the formation of a quinonediazide. The decomposition of the latter in light causes the solution to bezome deeply coloured.The solid product of the reaction which remains in solution in the methyl alcohol was obtained by evaporating the solvent after decomposition had been completed, sodium carbonate having been added to neutralise the free sulphuric acid. The solid residue, which was highly coloured, was distilled in steam ; from the colourless solid distillate a small quantity of s-tribromophenol was extracted by sodium hydroxide. It melted a t the correct melting point, 95'. The main portion of the solid, which was insoluble in sodium hydroxide, melted a t 60-65'. It was thought that this solid was a mixture of s-tribromobenzene (m.p. 120') and s-tribromoanisole (m. p. 88'), but crystallisation from alcohol did not change the melting point.* It was therefore treated with hydriodic acid in order t o convert the ether into s-tribromophenol. Fmm the product both a-tribromobenzene and s-tribromophenol were easily isolated, no indications of the presence of a third substance being noticed. The relative proportions of these two substances were determined by estimating the methoxy-group and determining the percentage of bromine in the mixture. The two methods agreed closely, within one per cent., the mixture containing 63-61, per cent. of the ether. The methoxy-group was estimated by Zeisel's method, using a mixture of acetic anhydride and hydriodic acid to decompose the ether; two hours' heating at 160' was found necessary for complete decomposition.Efeect of the Presewe of Water or Acids.-Addition of acid or dilution of the methyl alcohol with water does not greatly change the rate of the decomposition in sunlight ; in both cases there is a slight decrease. In the presence of acid (15 per cent. H,SO,), bromine is * The solubilities of s-tribromobenzene and s-tribromoanisole are very siniilar in both methyl and ethyl alcohols. Both are moderately soluble in boiling 90 per cent, ethyl alcohol. At 15"' 100 C.C. of this alcohol dissolves 1'03 grams of the tribromo- anisole, and 0'4 gram of the tribromobenzene ; these are almost the proportions in which these two substances are found in the product of decomposition of the di- azonium salt in methyl alcoholic solution.48 ORTON, COATES, AND BURDETT: THE INFLUENCE OF not eliminited, and in consequence no colour, owing to the decom- position of quinonediazide, developed.The proportion of s-tribromobenzene and s-tribromoanisole is con- siderably modified in favour of the former, which now forms the main product of the reaction ; s-tribromobenzene separates in the pure state during the decomposition. Owing to this effect of acid, high concentrations of the diazonium hydrogen sulphate, 4 per cent. and upwards, yield mainly s-tribromobenzene on exposure to light. I n such a solution, the concentration of the sulphuric acid is initially decinormal, and as the decomposition proceeds becomes fifth - normal, Behaviour of A Zka Zidiaxo t ale in Met hp? A Icoho Z ic XoZut iolz.-T ho behaviour of methyl alcoholic solutions of sodium or potassium s-tri- bromobenzenediazotates offers st marked contrast to that of the diazonium salts.A 1 per cent, solution of potassium diazotate remained quite colourless and unchanged during four hours' exposure to light. Decomposition of s-~ribromobenxenediaxoniu~~ liydrogen SulplLate in Et?~yll AZcoltoZic 8oZution.-s-Tribromobenzenediazonium hydrogen sulphate is very sparingly soluble in ethyl alcohol at the ordinary temperature, 100 C.C. dissolving somewhat less than 0.2 gram. I n 90 per cent. alcohol, on the other hand, the salt dissolves freely. A suspension of the salt in absolute alcohol changes very slowly at the ordinary temperature when protected from the light, the amount of nitrogen evolved in twenty-four hours being only just detectable.On the other hand, the solution of the salt in 90 per cent. alcohol is unstable even in the dark; a 1 per cent. solution continuously evolves nitrogen (temperature, 11-12'>, 50 per cent. of the salt being decomposed in twenty-four hours. s-Tribromobenzene is the main product of the change, but a t the same time bromine is eliminated, representing about 10 per cent. of one atomic proportion. On exposure to sunlight, both the suspension of the salt in absolute alcohol and the solution (1 per cent.) in 90 per cent. alcohol decompose very rapidly. Whilst the 90 per cent. alcohol becomes very deeply coloured, the absolute alcohol assumes only a pale yellow tint. The solid products in the two cases do not materially differ, and consisted of s-tribromophenetole and s-tribromobenzene.Dilution of the alcohol with water retards the decomposition ; thus two similar solutions in 50 per cent. and 25 per cent. alcohol had decomposed respectively to the extent of 82 and 62 per cent. after exposure to diffused daylight for five and a half hours. I n both cases, s-tribromobenzene was the main product, a small amount of bromine being eliminated. The effect of the presence of excess of acid was tested byLIGHT ON DIAZO-REACTIONS. I. 49 exposing a 1 per cent. solution of the diazonium szlt in a 15 per cent. solution of sulphuric acid in 90 per cent. alcohol. The decomposition took place rapidly, the solution acquiring only a pale yellow tint.No bromine was eliminated, and therefore no quinonediazide formed, a result in accord with a11 previous observations as to the effect of acid on the decomposition of this diazonium salt. Co-related with the non- formation of the quinonediazide is the absence of colour. It is note- worthy that in this case also s-tribromobenzene is the only product of the reaction. The foregoing observations indicate that the most favourable conditions for the formation of the phenetole are to be found in the use of pure alcohol and the absence of acid. The preparation in quantity was accordingly carried out by exposing a suspension of the diazonium salt (4 grams) in 100 C.C. of alcohol, When all the solid had dissolved and the evolution of nitrogen had ceased, the solution mas evaporated to half its volume, whereupon pure s-tribromobenzene (m.p. 1 19-120°, 1.2 grams) separated. Since s-tribromobenzene is but slightly soluble in cold alcohol, almost the whole of this substance is thus found. The mother liquor was then evaporated to dryness and the product recrgstallised from alcohol, but such treatment failed to raise the melting point above 60-65 ', whereas s-tribromophenetole melts a t 74". Control experiments, moreover, showed that it mas difficult t o isolate from mixtures of s-tribrornobenzene and the phenetole, and then only with great loss, any pure phenetolc. The composition of the mixture was accordingly ascertained from estimations of bromine and the ethoxy group ; * both analyses agreed in showing that the mixture contained 70-72 per cent.of the ether. After treating the mixture with hydriodic acid, the resulting s-tribromophenol and s-tribromo- benzene could easily be separated by alkali. From these results it follows that the original mixture of s-tribromobenzene and s-tribromo- phenetole contained about 5.7 per cent. of the former. Decomposition of s-l'ribromobenzenediazonium Ifiidrogen Sulphate in Acetic Acid Solution.-s-Tribromobenzenediazonium hydrogen sulphate dissolves freely in glacial acelic acid (99 per cent.) ; the solution is quite stable in the dark at the ordinary temperature, but becomes coloured on boiling, s-tribromobenzene being formed. On exposing a solution (1 per cent.) t o sunlight, decomposition is rapid and complete; the addition of an equal volume of water causes.the ;y I n order to obtain an accurate estimation of the ethoxy-group in s-tribromo- yheiietole by Zeisel's method, very prolonged heating of the substance with hydr- iodic acid (1 vol.) and acetic aiiliydride (0-5 vol.) is necessary. Thus, in one deterniiiiation (s-tribromophenetole = 0.3426 grain), ethyl iodide only ceased to be given off after six hours' heating, the percentage of ethoxy-group being found = 11 '54, whilst C,H,Br,*OEt requires 12 -19 per cent. VOL. XCI. E50 ORTON, COATES, AND BURDETT: THE INFLUENCE OF separation of colourless needles which melt a t 8 2 O , the melting point of s-trihromophenyl acetate, C,H2Br;OAc. An analysis showed that this substance was quite pure : 0.2103 gave 0.3156 AgBr. Er = 64.4G. The effect of diluting the acetic acid or of adding sulphuric acid is t o decrease the rapidity of the decomposition, water having the greater influence.The solid product obtained from the dilute acetic acid consisted of a mixture of s-tribromoplienol and its acetyl derivative in about equal proportions, A small amount of phenol was also present in the product obtained from the mixture of acetic nnd sul- phuric acids. C,H,O,Br, requires Br = 64.33 per cent. Decomposition of s-Fribrornobe?~xenedic~xonium Hyclyogen XuZp?mte in Various Media. POY~ZZ'C Acid-The diazonium salt is readily soluble in formic acid (90 per cent.); in the dark, the solution was shble, no gas being evolved. On exposure to sunlight, a very rapid decomposition set in ; the solid product consisted mainly of s-tribromobenzene mixed with a small amount of s-tribromophenol ; no phenyl Eormate was discovered.Since the formic acid had acted as a reducing agent, carbon dioxide should be present in the evolved gas. Examination of the gas showed this supposition t o be correct. Propionic Acid.-The diazonium hydrogen sulphate does not per- ceptibly dissolve in propionic acid. When the suspension is exposed to sunlight nitrogen is slowly evolved, and the solution becomes coloured a deep yellow. Neither s-tribromobenzene nor s-tribromo- phenol are formed, but owing to the slowness of the decomposition sufficient material could not be accumulated to determine its identity with tribromophenyl propionate. SuZphuric Acid.-The diazonium salt is readily soluble in concen- trated sulphuric acid (95 per cent.).The colourless solution slowly evolves nitrogen in sunlight, On cautiously diluting the solution, pure s-tribromophenol crystrallises 0 1 7 t free from any trace of by- product. rli'itric Acid.-Fuming nitric acid (sp. gr. 1.5) dissolves the diazonium salt freely. The decomposition in sunlight is complicated by the concomitant decomposition of the nitric acid, and by the interaction of the acid with the primary product ,which is probably s-tribromophonol. Bromine is partly displaced by the nitro-group, and can be estimated in the solution. It was not possible to isolate a single substance from the mixture of nitrobromophenols, but picric acid appeared not to be present.LIGHT ON DIAZO-REACTIONS. I. 51 Organic Lipuids.-The diazonium salt does not dissolve to any extent in glycerol, lactic acid, butyric acid, or ethyl acetoacetate.The suspension in the two solvents first named decomposes fairly rapidly, s-tribromobenzene being formed in each case. I n the other two media, the decomposition is very slow. Decomposition of Other Substituted Biazobenxenes. 2 : 6-Dibromodiaxobenxene.-The behaviour of this diazonium hydro- gen sulphate towards light is similar to that of the s-tribroino-deriv- ative, and affords an extremely easy method of preparing 2 : 6-dibromo- phenol. In aqueous solution the decomposition is very rapid, a very small quantity of bromine (2.54 per cent. of one atomic proportion) being eliminated. The phenol was freed from a small quantity of brown solid, the decomposition product of the bromoquinonediazide, by dissolving in sodium hydroxide. The phenol melted a t 56’ and gave the following numbers on analysis : 0.147 gave 0,2186 RgBr.Br = 63.3. C,H40Br2 requires Br = 63.5 per cent. This phenol can also be prepared easily from the diazonium salt by Heinichen’s method (Zoc. cit.), using a 63 per cent, solution of sul- phuric acid. Decomposition is complete in three-quarters of an hour ; the yield is good and no appreciable amount of quinonediazide or other substance seems to be formed. 2 : 3 : 4 : 6-Tetrabromodic~xobenxe7ze.-In the decomposition of aqueous solutions of this diazonium hydrogen sulphate by sunlight, the quinonediazide decomposition takes a more prominent place, some 15 per cent. of one atomic proportion of bromine being eliminated.The tetrabromophenol which is mainly formed can be obtained in a purer state when the decomposition is effected in 15 per cent. sulphuric acid, the quinonediazide decomposition being then inconsiderable. The phenol melted a t 115’ : 0.1448 gave 0,2645 AgBr. Br = ‘77.7. When this salt was heated in 63 per cent. sulphuric acid solution, a rapid and complex decomposition took place. As much as 53.6 per cent. of one atomic proportion of bromine appears in the solution, Very small amounts of two crystalline products were noticed, but no tetrabromo2henol was isolated. 2 : 4 : 6-l’ribrorno-3-nitrodicmobelzxene.-The decomposition of this diazonium salt, which is very rapid when a a per cent. solution is exposed t o sunlight, appears t o be very complex ; 46-47 per cent.of one atomic proportion of bromine is eliminated, and the solid product C,H20Br4 requires Br = 78.01 per cent. E 252 ORTON, COATES, AND BURDETT: THE INFLUENCE OF is a mixture of the decomposition product of the quinonediazide and the phenol arising directly .from the diazonium salt. The addition of sulphuric acid reduced the proportion of the quinonediazide decom- position ; nevertheless some 17.8 per cent. of bromine was eliminated. p-n~itrodi~zobenxene.-"he existence of the remarkably stable y-nitrobenzenediazotates and the corresponding nitrosoamine has rendered possible the testing of the action of light on these classes of substances. A dilute aqueous solution of the pure sodium p-nitro- benzenediazotate was exposed to light for several days. The liquid darkened slightly, and about 2 C.C.of gas were evolved. Examination of the solution after exposure showed that the diazotate was mainly unchanged. p-NitropT~enyZnitrosou~nine (pnitrobenzenediazohydroxide), which was prepared by Schraube and Schmidt's method from sodium p-nitro- benzenediazotate, is insoluble in cold water and slowly decomposes at the ordinary temperature. In the dark a t 0' it is, however, quite stable. A suspension in water was exposed to sunlight in a flask which was immersed in ice and water, the light having access to one side of the flask. The substance rapidly darkened and nitrogen was slowly evolved, 4 C.C. being collected during two hours' exposure. Decomposition of Diaxobenaene in Light. Inasmuch as acid solutions of diazobenzene readily undergo the phenolic decomposition a t the ordinary temperature, testing of the accelerating effect of sunlight on this change was somewhat difficult, Moreover, sunlight produces other changes i n solutions of this diazonium salt, which result in causing the liquid to become dark and turbid, effects which prevent light from gaining access i o the liquid and thus stop the reaction. Two exactly similar solutions of the diazonium hydrogen sulphate were made up in 20 per cent.sulphuric acid, and placed in flasks connected with nitrometers. One was exposed to sunlight and the other placed in a dark room, each being immersed in mixtures of ice and water. The solution, which was exposed to sunlight, became dark and turbid, and nitrogen was slowly evolved.The solution in the dark did not change in appearance, but a small volume of nitrogen was evolved, at about a quarter of the rate of the solution exposed to light. The solutions, still thoroughly cold, were extracted with chloroform in order to isolate any phenol. The residues left after evaporating the chloroform were distilled in steam, and the aqueous solutions treated with bromine water. The distillate from the solution which was exposed to light gave a precipitate of s-tribromophenol (m. p. 95"), whilst that from the screened solution gave only a slight turbidity.LIGHT ON DIAZO-REACTIONS. I. 53 Similar experiments were made with a '' normal I' diazo-solution, which was made up by adding 25 C.C. of a solution of benzenediazonium chloride (from 1 gram of aniline) to 30 C.C. of a 10 per cent.solution of sodium hydroxide. This solution decomposes when kept in the dark a t the ordinary temperature, 25 C.C. of nitrogen being evolved in twentyfour hours. A similar solution, in which, however, 30 C.C. of a 30 per cent. solution of sodium hydroxide were used, instead of the 10 per cent. solution, was more stable, and decomposed very slowly in the dark; light did not appear appreciably to hasten the rate of decomposition. Decomposition of illeth~ldiaxobenxenes. 5-B~*o.rlzo-4-diaxo-m-xyZene.-TTnlike other s-trisubstituted anilines in which negative groups such as the nitro-group or bromine are present, 5-brorno-m-xylidine can be diazotised in dilute acid solution. The solid diazonium salts can be prepared very readily in the usual way.Aqueous solutions of the diazonium salts are quite stable a t the ordinary temperature, provided that they are screened from light. Thus a 1 per cent. solution of the diazonium hydrogen sulphate in 15 per cent. sulphuric acid was kept in a flask connected with a nitro- meter for seventy-three days, the temperature ranging from 10-15O during that period. No gas was evolved, but when this solution was exposed to sunlight, a rapid decomposition set in, gas being given off, and an oil separating. No bromine is eliminated and the diazonium salt can be converted quantitatively into the xylenol. This change can be brought about with equal readiness by boiling the aqueous solutions of the diazonium salts. Very different is the behaviour of solutions of the diazotates.I n the dark they are quite stable, and when exposed to sunlight only show signs of decomposition if the excess of alkali is slight and consequent,ly the hydrolysis considerable. , - 5-Byorno-m-4-x yleno I, BI-, )o M * OH This xylenol appears to have been prepared by Noelting, Braun, and Thesmar (Bey., 1901, 34, 2242) from the corresponding xylidine. These authors state that the xylenol is a solid melting a t 72'. We have, however, found that the xylenol obtained from 5-bromo- m-4-xylidine by diazotising and then decomposing the diazonium salt either by boiling or by exposing i t s aqueous solution to light, is an oil a t the ordinary temperature. Moreover, bromination of m-4-xylenol54 ORTON, COATES, AND BURDETT: THE INFLUEKCE OF also yields a monobromo-derivative which is identical with the substance obtained from the bromoxylidine.Bromifiation, of rn-Li-Xylenol.-A solution of bromine (6.55 grams) in acetic acid was added to a solution of m-4-xylenol(5 grams) in the same solvent in which fused sodium acetate (3.6 grams) was suspended. Bromination was instantaneous, heat being developed. The xylenol was separated by addition of water, and distilled in steam, and after separation from the aqueous distillate by chloroform and drying in the latter solvent, the oil was fractionated under reduced pressure. It was colourless, boiled a t 110.5" under 18 mm. pressure, and had a sp. gr. 1.4569 at 11'/4" : 0.1235 gave 0.1141 AgBr. Br = 39.89. The bromoxylenol prepared from the bromoxylidine by way of the diazo-compound, boiled at 112" under 20 mm.pressure, and had a sp. gr. 1.4607 a t 11"/4". An analysis of a specimen obtained by boiling a solution of the diazonium salt gave the following numbers : 0.1215 gave 0.1145 AgBr. A specimen obtained by exposing a solution of the diazonium salt to light gave the numbers : C,H90Br requires Br = 39.80 per cent. Br = 40.10 per cent. 0.1098 gave 0.1 018 AgBr. Br = 39.45. C8H90Br requires Br = 39.80 per cent. B-Bromo-$-cumidine (6-Bromo-5-ccmino-1 : 2 : 4-trimetl~yZbenxene).- This base has apparently hitherto not been described. It is readily prepared by brominatiog $-cumidine in acetic acid solution in the presence of sodium acetate, care being taken to add the solution of the base t o the solution of the bromine. The bromo-+-cumidine is precipitated by addition of water and purified by distilling in steam.It crystallises from dilute alcohol in long, colourless needles melting at 68-69' : 0.1906 gave 0.1675 AgBr. Br= 37.4. The salts of this base are not markedly hydrolysed by water, thus resembling those of bromoxylidine. It is very readily acetylated by heating for a short time with acetic anhydride a t 100'; the acetyl derivative crystallises in clusters of four-sided prisms melting at 206'. C,H,,NBr requires Br = 37.4 per cent.LIGHT ON DIAZO-REACTIONS. 1. 55 It can easily be diazotiscd in dilute acid solution, and the solid diazonium hydrogen sulphate can be prepared in the usual way. Acid solutions of bromo-+-diazocumene behave in exactly the same way as do those of bromodiazoxylene.Kept in the dark at the ordinary temperature they are quite stable, but on exposure to light, gas is immediately evolved, and a turbidity soon appears in the solution. Pure 6-bromocumenol (m. p. 32 -33') slowly separates in needles. Boiling of the aqueous solutions of the salts brings about a rapid and quantitative conversion into the bromocumenol. 5-Dinxo-+-cun~ene.-An attempt was made to contrast the de- composition of an acid solution of diazo-$-cumene exposed to light with one kept in the dark. A t the ordinary temperature, this diazo- compound decomposes fairly quickly, whether screened from light or not. A t 0-2', a marked difference is observed ; whilst the solution which is kept in the dark very slowly evolves nitrogen, the other decomposes rapidly. +-Cuinenol crystallised out as the decomposition proceeded, and melted a t 71". 3 : 5-Dibronao-o-dic~xotoluene, and 3 :~5-Di6rorno-p-diuxotoZuelze.-Th~ dibromodiazotoluenes resemble very closely s-tribromodiazobenzene. Solutions of the hydrogen sulphates are quite stable in the dark, but on exposure to light rapidly decompose. I n aqueous solution, a small amount of bromine is eliminated, and the 3 : 5-dibromocresol, which separates during the change, is discoloured by the products of de- composition of the quinonediazide. Quantitative experiments have shown, however, that only 10 per cent. of the diazonium salt under- goes the quinonediazicle decomposition, the remaining 90 per cent. being converted into cresol. If the decomposition is carried out in 15 per cent. sulphuric acid solution, no bromine is eliminated, and the cresol separates in almost pui e, colourless needles. The cresols can easily be isolated from the product of the action of light by distilling in steam. 3 : 5-Dibromo-o-cresol crystallised in long needles melting a t 55' : 0.13'72 gave 0.1932 AgBr. 3 : 5-Dibromo-p-cresol was obtained in needles melting a t 43-4 1" : 0.1506 gave 0 2123 AgBr. When dilute (1 pcr cent.) aqueous solutions of the sulphates are heated a t loo', rapid decomposition ensues; about 70 per cent. of the total nitrogen in the diazo-compound is given off a s gas. A dark brown, amorphous solid separated from which no cresol could be isolated; the filtrate was quite colourless and free from both un- changed diazonium compound or quinonediazide, but a trace of bromine mas present. Br = 59.93. C7H,0Br2 requires Br = 60.1 2 per cent. Br = 59.98 per cent.56 BERKELEY: ON THE MORE EXACT DETERMINATION OF The dibromodiazotoluenes differ from tribromodiazobenzene in that they readily yield the corresponding cresols when they are heated i n solution in 63 per cent. sulphuric acid. Decomposition is very rapid, and a crystalline substance distils over, which is mainly the cresol. Solutions of the sodium dibromotoluenediazotates can be easily prepared, and accordingly offer an excellent means of demonstrating the stability of these substances t o light, After three days’ exposure of the sodium derivative of dibromo-o-diazotoluene in 1 per cent. aqueous solution, the liquid was still quite colourless ; no measurable amount of gas had been evolved. The authors are indebted to the Chemical Society and to the British Association for the Grants which have largely defrayed tho cost of these researches. They wish to take this opportunity of expressing their thanks to these Societies. BANGOR. UNIVERSITY COLLEGE OF NORTH WALES,