XXIL-Aronmtic Compounds Obtained from the Hycho-aromatic Series. Part III. Bromoxylenols fromDirnethy ldihydy-o resorcin.By ARTHUR WILLIAM CROSSLEY and NORA RENOUF.IN the course of an investigation on the action of phosphorushaloids on diinethyldihydroresorcin (Crossley and Le Sueur, T.,1903, 83, llO), it was shown that the primary products of thereactions, which were of a hydroaromatic nature, readily underwentrearrangement t o I orm aromatic substances. More particularly wasthis the case when phosphorus pentabromide was employed, andfrom the resulting products there were isolated a monobromoxylenol,melting a t 84O, and a dibromoxylenol, melting a t 9To, both of whic166 CROSSLEY AND RENOVF : AROMATIC COMPOUNDSsubstaiices wer.e showii t o be derivatives of 0-3-xylenol, because 011treatment with bromine they gave tribromo-o-3-xylenol, melting at183O.Further, it was stated that a second tribromoxylenol hadbeen encountered, melting a t 177O, but this ha.e zow b&n proved,despite its constant melting point, t o be a mixture of the tribromo-derivatives of o-3-xylenol and o-4-xylenol.The question of the constitution of the mono- and di-bromo-xylenols was not studied in detail at the time, although it waspointed out that the conversion of the hydroaromatic substancesinto aromatic compounds took place so readily, that the reactionsmight afford the easiest means of preparing certain substitutedxylenols, otherwise difficult of production. Such is indeed the case,as is proved by the work now described.On renewing the study of the direct action of phosphorus penta-bromide on dimethyldihydroresorcin, it was soon found that thereactions are of an extremely complicated nature, being susceptibleto the slightest variation in experimental conditions.For example,on repeating the experiment (Zoc. cit., p. 128) in which the mono-bromoxylenol, melting a t 84O, was originally obtained, and employ-ing apparently precisely the same conditions, neither the mono-bromoxylenol (m. p. 84O) nor the dibromoxylenol (m. p. 9 7 O ) wasactually isolated, but in their place a new bromoxylenol, melting a t103O. It is, however, certain that, with the additional knowledgeof the properties of the derivatives of these substaiices, accumulatedduring the progress of tlie present work, all three xylenols coiildnow be isolated froin the mixture.I n the paper already alluded to, it was shown that the primaryproducts of the action of phosplioru s pentabromide on dimethyl-dihydroresorcin or of phosphorus tribromide on bromodimethyl-dihydroresorcin were dibromo- and tribromo-dimethylcyclohexenones(I and II), and the statement was made that “the xylenol obtainedin the latter reaction must be produced by rearrangement of eithermono- or di-brornodimethylcyclohexefiones (dibromoketodimetE,y!-tetrahydrobenzene) .’,H,C{ \CH2ErC\\,CO(1.) (11.1It was therefore decided to study the possible transformationsof these substances and also tribromodimethylcyclohexenone toaromatic compounds, because, as their constitutions are known, iOBTAINED FROM THE HYDROAROMATIC SERIES.PART III. 167would be easier to obtain clues as to the orientation of the bromo-xylenols produced from them.Tho reagents employed for the transformations were potassiumhydroxide in alcoholic solution, which gives rise to derivatives ofo-3-xylenol only ; and heat, which produces derivatives of botho-3-xylenol and o-4-xylenol.Other reagents were also tried, for example, (a) sulphuric acid,which, however, is of little value, because, although xylenols areproduced, the reaction is complicated by the presence of bromine,due t o the action of the sulphuric acid on the liberated hydrogenbromide, which appears to cause bromination to take place in theside-chain as well m in the nucleus; ( b ) nitric acid (D 1-42}, whichdoes not give rise-to aromatic compounds, but partly oxidises thehydroaromatic bromoketones, and also brominates them (compareT., 1904, 85, 273); ( c ) diethylaniline, which does not transformthe hydroaromatic substances when heated to 210-220° for a fewminutes, and on prolonged heating gives rise only to 'resinousproducts.The various bromoxylenols isolated in the course of the workwere identified, although as a rule no details are mentioned in thepractical portion of the paper, by analysis in some cases, andalways by the determination of the mixed melting points of thesubstances themselves and of their benzoyl derivatives with syntheticpreparations of similar compounds.Having first established the fact that bromodimethylcyclo-hexenone does not give aromatic compounds with the above-mentioned reagents, the next action tried was that of potassiumhydroxide on dibromodimethylcyclohexenone (111), which givesrise to ti-bromo-o-3-xylenol (IV) and small quantities of 4 : 5-di-(IV.) (111.) (V.)bromo-o-3-xylenol (V). The yield of bromoxylenols is not large,and the main product of the reaction is a liquid, insoluble inpotassium hydroxide, which appears to consist mainly of ethoxy-compounds, their formation being explained by the fact that thereaction is carried out in alcoholic solution. The formation ofethoxy-compounds under these conditions has frequently beennoticed, as, for example, in the action of alcoholic potassiumhydroxide on 1 : 2-dibromocycZohexane, when the product is mainlethoxycyclohexene (T., 1904, 85, 1415; compare also T., 1905, 87,1499).The above rearrangement is extremely interesting, but of asomewhat startling nature when an attempt is made to account inan adequate manner for its mechanism.No simple explanationis forthcoming, for the reaction necessitates a wholesale migrationof atoms of which the most unexpected is perhaps the initialremoval of hydrogen bromide, the bromine from carbon atom 4with a hydrogen atom in the meta-position to it a t either 2 or 6.Simultaneously with this movement a hydrogen atom in eitherposition 2 or 6 must wander into the meta-position to replace tliebromine atom thus removed.From a casual glance at the formula for dibroniodimethylt.yc.lo-hexenone, i t would seem that tlie bromine atom in position 5 woul(1certainly be tlie one to be removed as hydrogen bromide togetherwith one of tlie hydrogen atonis in position 6.Yet such is iiotthe case, for the constitution of the resulting bromoxylenol hasbeen proved by synthesis (T., 1913, 103, 2179), and therefore therecan be no doubt that it is the bromine atom in position 4 which iseliminated.It is, however, t o be noted that, as in so many instances alreadyquoted (T., 1902, 81, 1533; 1904, 85, 264; 1906, 80, 875; 1908,93, 633), the methyl group has again wandered into an ortho-position.It would seem bhat the production of a small amount of 4: 5-di-bromo-0-3-xylenol in this process can only be due to side reactions,where bromine is liberated, causing further bromination, of5-bromo-o-3-xyleno1, as it is impossible for a dibromoxylenol toresult from dibromodimethylcyclohexenone by the removal of theelemenb of hydrogen bromide.When dibroniodimethylcyclohexenone is heated, hydrogenbromide is evolved, and about half the weight of substance takenis recovered as a mixture of 5-bromo-o-3-x~lenol (VI), melting a t84O, and 6-bromo-o-4-xylenol (VII), meltingCH3 C(CW2/ b H 3 H,Cf\C H2B ~ , ,!OH BrC\)CO(TI.) CBr (VII.)There is again the same difficulty iii offering any adequateexplanation of these transformations, but a new point of interestis raised.Again a methyl group has wandered into an ortho-position, but in the one case from position 1 to position 2, whereasin the second case, the wandering is from position 1 to position 6ORTAINET, FROM TElE H\’DlIOAROMATIC! SERIES.PART 111. 169This is the first occasion on which this particular wandering hasbeen noticed by the present authors.Although, as already pointed out, it would seem probable thatthe hydrogen and bromine atoms in positions 5 and 6 would beeliminated as hydrogen bromide, this is not the case, for hydrogenbromide is again removed from carbon atoms in the meta-positionto one another. The produdion of derivatives of 0-3-xylenol andof o-4-xylenol is due to the swinging of a methyl group in theone case from position 1 t o position 2, and in the other casefrom position 1 to position 6, otherwise the rearrangement is of anentirely similar nature.In the previous paper (loc.cit., p. 115) it was stated tJiat broino-dimethylcyclohexenone readily absorbed bromine, giving offhydrogen bromide and undergoing a transformation which had notbeen worked out,. so that dibromodiruethylcyclohexenone (VIII)could not be obtained by the action of bromine on bromodimethyl-cyclohexenone (IX), which seemed somewhat remarkable in viewof the action of bromine on the similarly constituted dimethyl-dihydroresorciii and of the stability of dibromodimethylcy d o -hexenone.(VIII.) (IX.1I n again carrying out this action of bromine on bromodimethyl-cyclohexenone, although only minute quantities of dibromodimethyl-cyclohexenone have been actually isolated, there can be no doubtthat this substance is, in reality, the main product of the reaction,although it cannot be separated on account of admixture withunchanged bromodimethylcycEohexenone and some tribromodi-methylcyclohexenone, as is also the case in the action of phosphorustribromide on bromodimethyldihydroresorcin (Zoc. cit., p.121).This would seem t o be demonstrated beyond doubt by a study ofthe transformation products of the liquid obtained initially fromthe action of bromine on bromodimethylcyclohexenone, which istotally insoluble in potassium hydroxide, and entirely hydro-aromatic in nature.This crude material gave, on treatment with potassium hydroxidein alcoholic solution, mainly 5-bromo-o-3-xylenol and smallquantities of 4 : 5-dibromo-o-3-xyleno1, in approximately the sameproportions as when these xylenols were obtained by the actionof alcoholic pot,assium hydroxide on pure dibromodimeth ylcpclo-1 texenone170 CROSSLEI' ANT) RENOUF : AROMATIC COMPOUNDSMoreover, the crude product ~ gave, when heatled, 5-bromo-o-3-xylenol and 6-bromo-o-4-xylenol, again the same products andin approximately the same proportions as have been obtained bythe action of heat on pure dibromodimethylcyclohexenone.Incidentally, it may be mentioned that this process provides amuch easier method for the preparation of these two bromoxylenolsin quantity than the methods described for their syntheses (T., 1913,103, 1297, 2179).When tribromodimethylcyclohexenone (X) is transformed eitherunder the influence of heat or of potassium hydroxide in alcoholicsolution, the only bromoxylenol isolated is 4 : 5-dibromo-o-3-xyleno1,melting a t 97O.There is a doubt as to the position of the thirdbromine atom in tribromodimethylcydohexenone (Zoc. cit., p. 114),which may be represented by either of the two formulze X or XI,(X. 1 (XI.)although the former would appear to be the more likely. Which-ever of the two positions is actually occupied by the bromine atomis, however, of no importance on the present occasion, because, asshown by the synthesis of 4 : 5-dibromo-o-3-xylenol (T., 1913, 103,989), this is the bromine atom eliminated from the molecule, andhence again hydrogen and bromine are removed as hydrogenbromide when occupying the meta-position to one another.EXPER TMENTAL.Preparation of 4 : 5-Dibromo-1 : l-dimethy2cyclohexen-3-one(4 : 5-Da'brom0-3-keto-l: l-dimet~~l-L\4-tetrahy~r~~enzene).The conditions originally given (T., 1903, 83, 121) for the p r eparation of dibromodimethylcyclohexenone have been modified, thequantities of materials now used being 25 grams of anhydrousbromodimethyldihydroresorcin (3 mols.), 100 grams of dry chloro-form, and 15.5 grams (14 mols.) of phosphorus tribromide, andheating was only continued for one and a-half hours.The yieldof dibromodimethylcyclohexenone is thereby doubled, being from6 to 7 grams, but although the mother liquor ( A , see below) containsmuch more of this substance, itl cannot be isolated on account ofits solublity in the accompanying bromodimethylcyclohexenone.In the previous paper (loc.cit., p. 114) it was pointed out that(( the production of bromodimethylcyclohexenone (XII) from bromoOBTAINED FROM THE HYDROAROMATIC SERIES. PART IIT. 171diinethyldihydroresorcin (XIII) is a reaction for which no adequateexplanation is at presoiit forthcoming.”C(CH,\,H,C/\CH,ROC1 ICO \/CBr(XII.) (XIII.)It is, however, most probably due t o the fact that bromodimethyl-dihydroresorcin crystallism with 1H,O, which is not easy to removecompletely when .working with the substance in large bulk. Hencethe hydrogen bromide formed by the action of phosphorus tribromidewith this molecule of water, would hydrolyse the bromodimethyl-dihydroresorcin to dimethyldihydroresorcin, which would then beacted on by phosphorus tribromide to form bromodimethylcyclo-hexenone.This supposition seems to be supported by the fact thatin one preparation where cryst,alline bromodimethyldihydroresorcinhad been employed, without being dried in a vacuum, the resultingmaterial consisted principally of bromodimethylcycZohexenone, or,a t all events, no dibromodimethylcyclohexenone crystallised out.Action of Potassium Hydroxide on Dibromodimetkylcyclohexenone.Five grams of dibromodimethylcyclohexenone were dissolved in200 C.C. of absolute alcohol, and N / 4-alcoholic potassium hydroxidegradually added in small portions, and the whole heated, betweeneach addition, until no longer alkaline. The amount of potassiumhydroxide used was 0.994 Ram, corresponding exactly with onemolecule.The major portion of the alcohol was evaporated, thewhole poured into water, acidified, extracted with ether, theethereal solution washed with potassium hydroxide solution, thewashings acidified, and distilled in a current of steam. On extra&ing the distillate with ether, etc., a residue of 1 gram remained,which was recrystallised from light petroleum (b. p. 40-60°), when0.6 gram of radiating clusters of long, transparent needles, meltinga t 84O, separated, which gave a benzoyl derivative, melting a t 98O,neither of which melting pokts was altered on admixture withrespectively 5-bromo-o-3-xylenol and its benzoyl derivative.The light petroleum mother liquor was evaporated to dryness,the residue spread on porous plate, and crystallised from aqueousalcohol, when 0.1 gram of 4: 5-dibromo-o-3-xylenol (T., 1913, 103,989), melting a t 97O, was obtainedA ction of Iieat on D i h o moclirri e t Ji~lcyclolwxetm tLe.Five grams of dibromodimethylcyclohexenone were heated ona sand-bath in a flask attached to an air condenser until a reactionset in, when the source of heat was removed.The reaction wasvigorous, and torrents 0.f hydrogen bromide were evolved. Thewhole was dissolved in ether, washed with potassium hydroxide,the washings acidified and distilled in a current of steam, and thedistillate extracted with ether, when 2.3 grams of solid wereobtained. This was benzoylated and treated exactly as describedon p. 173, when there were obtained 0.7 gram of 5-bromo-o-3-xyleno1, melting at, 84O, and 0.5 grain of 6-?pino-o-4 -xyleiiol,melting at 103O.Many attempts have been made to extract larger quailtities ofthe dibromoketone froin the mother liquors 9 (see above), but,without success.It was originally investigated by submitting i tto fractional distillation, whgn it appeared to consist mainly ofbromodimethylqclohexenone Fnd a higher fraction of aromaticsubstances formed by the action of heat on the liyclroaromaticcompounds present.Under the altered conditions of preparation the mother liquorcontains not more than 25 per cent. of broinodimet hylcyclohexenone,together with nearly 70 per cent. of dibrornodimethylc~yc7ohexenone,as is proved by the following experiments.Twenty grams of the nmther liquor were heated and worked upas described above, under the action of heat on pure dibromo-dimethylcyclohexenone.There were obtained 6.3 grams of bromo-xylenols, from which 2.4 grams of pure 5-bromo-o-3-xylenol (m. p.84O) and 1.4 grams of 6-bromo-o-4-xylenol (m. p. 103O) wereisolated. Calculating from the amounts of these substances pro-duced by the action o f heat on pure dibromodimethylc!/cZo-hexenone, it would appear that nearly 14 of the 20 grams of motherliquor consisted of dibromodimethylcyclohexenone, which could notbe crystallised on account of its admixture with bromodimethyl-cyclohexenone. Moreover, similar results were obtained by theaction of potassium hydroxide on the mother liquor, when 5-bromo-0-3-xylenol and 4 : 5-dibromo-o-3-xyleno1 were obtained, in amountscorresponding with what would have been expected, on the assump-tion that the mother Iiquor contained about 70 per cent.ofdibrom odiinetliylcycEohexenone.When dibromdimethylcyclohexenone was warmed with concen-trated sulphuric acid, i t dissolved, and solid soon separated, which,after crystallisation from alcohol, melted a t 187-18807 but as thesubstance was n o t phenolic in natiire, it, was n o t further investi-gatedOBTAINED FROM THE HYDROAROMATfC SERIES. PART 111. 173A c tio rz of Bro inane o 15 Brom odim e $16 ylcyclohexenone.Bromodimethylcyclohexenone was prepared as previously de-scribed (T., 1903, 8 3 , 1 2 0 ) ; it boils a t 126O/32 mm., that is, some-what lower than stated, and the yield is about 40 per cent.of thetheoretical amount. The substance previously mentioned as meltinga t 296O was again encountered, but as it has now been shown tocontain phosphorus, i t is iiot proposed further to investigate itsiiat ur 6.Twenty grams of bromodimethylcycloliexenone were dissolved in40 grams of dry chloroform, and a solution of 16 grams of bromine(1 molecule), in an equal weight of dry chloroform, was graduallyadded, care being taken iiot to allow the temperature to rise aboveOo. No hydrogen bromide was given off, and the bromine was notcompletely absorbed.On removing from the cooliiig medium, the temperature verygradually rose to 1 5 O , then very rapidly to 28O, when a reactioiiset in, all the bromine was used up and hydrogen bromide evolvedin quantities.After evaporation of the chloroform, the residueweighed 29 grams, and was a pale yellow liquid, completely insolublein potassium hydroxide, and possessing a camphoraceous odour.On standing for months, only a very small amount of solidwas deposited, which proved to be a mixture of dibromo- and tri-bromo-dimethylcyclohexenones. As the liquid could not be distilledwithout losing hydrogen bromide with partial transformation intoaromatic substances it was decided to investigate the latter in orderto get an idea of its composition.In the first place, 12 grams of the crude product were treatedwith alcoholic potassium hydroxide exactly as described in the caseof dibromodimethylcyclohexenone (see p. 1 7 1 ), when, on extractingthe steam distillate with ether, 2 grams of solid were obtained,which yielded 1.0 gram of 5-bromo-o-3-xyleno1, melting a t 84O, and0.3 gram of 4 : 5-dibromo-o-3-xyleno1, melting a t 9 7 O .In the second place, 36 grams of the crude product were heatedin a flask attached to an air condenser and worked up in exactlythe same manner as described under the action of heat on dibromo-dimethylcy clohexenone, when the steam distillate yielded 17 gramsof solid bromoxylenols.Previous attempts to separate pure sub-stances from this mixture, by repeated fractional steam distillationand crystallisation from light petroleum, had only resulted in theisolation of a zery sinall ainount of 6-broino-o-4-xy1eno1, meltinga t 1 0 3 O ; but by adliering strictly t o the following conditions asharp separation can be effected.Seventeen grams of crude productwere benzoylated in the usual manner, and the resulting 25 gram174 CltOSSLEY AND BENOUF : AROMATIC COMPOUNDSof benzoyl derivatives dissolved in 500 C.C. of absolute alcohol,when, after standing, 5.8 grams of a substance melting a t 94-96Owere obtained. By slow evaporation of the alcohol further fractionsof similar melting point have sometimes been isolated, but as arule all fractions but the first melt between 65O and 75O. Oncrystallising this solid (m. p. 94-96O) from alcohol, the meltingpoint rose to 98O, and examination showed it to be the benzoylderivative of 5-bromo-o-3-xylenol (see T., 1913, 103, 2182). Thealcohol was then evaporated t o a low bulk, and the benzoylderivative hydrolysed by boiling with potassium hydroxide. Thealcohol was evaporated and the acidified product extracted withether, the ethereal solution washed with sodium carbonate solution,then with water, dried over calcium chloride, and evaporated, when13.2 grams of solid were obtained, which on crystallisation fromlight petroleum (b.p. SO-looo) gave 2.8 grams, melting a t 103O:0.1100 gave 0.1025 AgBr. Br=39.65.C8H,0Br requires Br = 39.80 per cent.The synthetic formation of this bromoxylenol, which proves it tobe 6-bromo-o-4-xyleno1, has already been described (T., 1913, 103,1297). After separation of this substance the light petroleummother liquor was evaporated, the residue benzoylated, and thewhole process repeated twice, when there were finally obtained,from the original 17 grams of bromoxylenols, 6 grams of 5-bromo-0-3-xyleno1, melting a t 8@, and 4.6 grams of 6-bromo-o4-xylenol,melting a t 103O.The numerous processes through which the inaterial has to betaken necessitate some loss, and finally a small residue of just overone gram was obtained, from which nothing further could beisolated. If any other bromoxylenol is produced during thereaction, other than those above mentioned, it can only be presentin very small amount.Preparation of 2 : 4 ; 5-Yribromo-1 : l-dimethylcyclol~,exe~~-3-otie(Tribromo ketodimethyltetrnh ydro b enzene) .The method of preparation of this substance (T., 1903, 83, 124)has been so much modified as to necessitate redescription.Twenty-eight grams of dimethyldihydroresorcin (one molecule)were suspended in 400 grams of dry chloroform, 32 grams of bromine(one molecule) gradually added, then 172 grains of phosphoruspentabromide (two molecules), and the whole heated 011 the water-bath for one hour.The major portion of the cliloroform was tlieiievaporated, and this is a point which considerably influelices theyield of material, which is diminished if it is attempted to removOBTAINED FROM THE HYDROAROMATIC SERIES. PART 111. 1’75the last traces of chloroform. The whole was then poured intowater, extracted with ether, the ethereal solution washed withpotassium hydroxide so”lution (washings =A), etc., when 54 gramsof a semi-solid mass were obtained, which after triturating withlight petroleum (b.p. 40-60°) gave 34 grams of a clean, whitesolid, and this, after crystallisation from light petroleum (b. p.80-looo), yielded 28 grams of pure tribromodimethylcyclohexenone,melting a t 107O.The potassium hydroxide washngs ( A , see above) were acidifiedand distilled in a current of steam, when there separated from thedistillate a small quantity of a bromoxylenol, crystallising fromdilute alcohol in glistening, transparent needles, melting a t 96-97O.It is not identical with 4: 5-dibromo-o-3-xyleno17 melting a t 97O,because the mixed melting point was 63O. The amount was, how-ever, too small for an investigation of its constitution, which mustbe left over for decision, until further experiments, now beingcarried out, are completed.Action of Potassium Hydroxide o n Tribromodimethylcyclohexenone.Ten grams of the tribromoketone were treated with alcoholicpotassium hydroxide solution exactly as described on p.171. Thesteam distillate yielded 6.9 grams of solid, which, after crys-tallisation from alcohoI, gave 1.3 grams of pure 4 : 5-dibromo-0-3-xylenol (T., 1913, 103, 989). A small quantity of some otherbromoxylenol is also produced in this reaction, but its identity hasnot so far been established.A c t io n of Een t on Tr i b r o modim e thy lcy clok ex e n on t .Tribromodimethylcy clohexenone was heated in quantities of5 grams at one time, as described on p. 172. It is essential for thesuccess of the experiment that the source of heat should be a t onceremoved on the appearance of the first signs of a reaction, asotherwise the whole mass is resinified. Twenty grams of the tri-bronioketone gave 9.5 grams of bromoxylenols, which, after treat-nient with light petroleum (b.p. 40-60°) in the cold, gave 2.4grams of 4 : 5-dibromo-0-3-xy~eno1, melting a t 97O.The portion of bromoxylenols soluble in light petroleum (6.5grams) is a mixture, which has defied, up to the present, all themany attempts that have been made to separate its constituents.The problem will again be attacked when a systematic series ofexperiments on the separation of various bromoxylenols, now incourse of progress, has been coppleted176 CROYSLBY ANb RENOUF ! AROMATIC! C'OllIPOtTNDS, ETC.Action of SuEphuric Acid on Tribromodimethylcyclohexenone.Although the results obtained from the action of sulphuric acidon tribromodimethylcyclohexenone are of an unsatisfactory nature,the following brief account is quoted.It shows that trans-forma.tions do take place, but also illustrates the difficulty, met within some cases, of deducing correct conclusions from the mixedmelting-point method for the identification of compounds.When tribromodimethylcyclohexenone is warmed with twentytimes its weight of concentrated sulphuric acid, i t readily passesinto solution, and almost at once needle-shaped crystals separate,2 grams from 3 grams of the tribromoketone. This solid containedbromine, was phenolic in nature, and after repeated crystallisationfrom alcohol, melted sharply a t 178--179O, nor was this ineltingpoint lowered on mixing with tribromo-rn-4-xylenol.Moreover, thesolid gave an acetyl derivative melting a t 120°, which is the meltingpoint of acetyl tribromo-m-4-xy1eno1, but on mixing with this lattersubstance, an appreciable lowering of melting point was observed(110O). The melting point was not lowered, however, on mixingwith acetyl tribromo-o-3-xylenol, so the substance was hydrolysed,and the bromoxylenol crystallised from alcohol, when it melted a t182--184O, nor was this melting point lowered on mixing with puretribromo-o-3-xylenol. Nevertheless the two substances are notidentical, for on again acetylating and fractionating the resultingacetyl derivative, the melting point of a small portion was finallyraised to 150--153O, which is a t least 25O higher than the meltingpoint of any known acetyl derivative of the various tribromo-xylenols.For the purposes of this research, it has been necessary to prepareseveral tribromoxylenols and their acetyl derivatives, because, 011consulting the literature, it was found that there were considerabledifferences in the recorded melting points f o r one and the samesubstance. Further, one of the acetyl derivatives had not beenpreviously described. There is therefore appended a list of themelting poinss of certain tribromoxylenols and their acetylderivatives, together with their mixed melting points, which mayprove of use to other investigators.The acetyl derivative of tribromo-o-3-xyleno1, prepared by heatingthe xylenol with acetic anhydride, crystallises from either glacial:icetic acid or alcohol in clusters of transparent needles, meltingat 120O:0.1054 gave 0.1484 AgBr.Br = 59.91.CloH,O,Br, requires Br = 59.85 per centTHE PO1,TSULPHIDES OF THE ALKALI METALS. PART 1. 177Melting Points of Tribromoxylenols and their Acetyl Derivatives.Tribromc-o-3-xylenol .. ....... 185" Acetyl derivative ............ 120Tribromo o-4-xylennl.. 172-173" ) ) ............ 112"Tribromo-~n.4-xylenol 178-179" ............ 121"Tribromo-p-2-xylenol 179-180" ............ 125"....... 9 99 9 ,,9 , ? I..... .....Mixed Melting Points of Tribromoxylenols and their correspondingAcetyl Derivatives.F. P.Tribronio-xyleiiols.Tribroino-o-3-xylenol+ tiib~ot1io-o-4-xylenol ...... 176-183"Tribromo-o-3-xylenol+ tribromo-m-4-xylenol .... 158-181"Tribromo-o-3-xylenol+ tribromop-2-xylenol ...... 180-1 82"Trihroi1~0-0-4-xylenol+ tribromo-m-4-xylenol.. .... 173-1 76"Tribromo-o-4-xylenol+ tribronio-p-2-xylenol ...... 177-1 78"Tiibrotno-nt-4-sylenol+ tribromo-p2-xylenol ...... 178-180"RESEARCH LABOBATORIES, PHARMACEUTICAL SOCIRTY,17, BI.OOMSBUBY SQUARE, W.C.31. p.Acetjlderivatives.110-113"109-1 10"121-121 -5"109-111"11 2-1 14"109-1 10