BENTLEP AND PERKIN: REDUCTION OF XYLIC ACID, ETC. 157 XIL-The reductiou of Xylic Acid, of Paraxylic Acid and of Methylterephth,aZic Acid, and the preparci- tion of Metlzylterephthulic Acid crnd of Methyl- isophthulic Acid. By WILLIAM HENRY BENTLEY and WILLIAM HENRY PERKIN, Jun. KACHLER (Annalelz, 1813, 169, 183) first showed that sulphocamphylic acid, U,H,,SO',, is decomposed when fused with potash with formation of a monobasic acid, C,H1,O,, melting at 148'. C9Hl,S05 = G,H,,O, + SO, + H,O,158 BENTLEY AND PERKIN : REDUCTION OF XYLIC, PARAXYLIC, and subsequently Damsky (Bey., 1887, 20, 2965) obtained, by the same reaction, a second acidof this formula melting a t 99O. This action of fused soda or potash on sulphocamphylic acid has been made the subject of an extended investigation by one of us (compare Proceedings, 1893, 109 ; 1895, 23 ; 1896, lS9), in the hope of discovering the con- stitution of the two isomeric acids C,H,,O,, and of thus obtaining evidence as to the constitution of sulphocamphylic acid and, indirectly, also, of that of camphoric acid itself.During the course of the examination of these two acids, which mere named a-camphylic acid (m. p. 14S0) and P-camphylic acid (m. p. about lo$"), some facts came to light which seemed to point t o the conclusion that these substances might be closely allied to xyZic ucid, I, or pa?*ccz~~~c COOH CH, acid, 11, and this seemed all the more probable when it was remembered that sulphocamphylic acid itself could be converted into derivatives of these acids by two widely different processes.When sulphocamphylic acid is oxidised with potassium permanganate (Proc., 1893, log), it yields a dibasic acid of the formula C,,H,,07, which, by the action of concentrated sulphuric acid, is converted into a hydroxyxylic acid of the probable coiistitution COOH 0 H o H 3 \/ CH3 Again, Koenigs and Meyer (Be?*., 1894, 2'7, isokc~u~onoZic acid, CgH1402, which is formed 3465) found that when by the action of heat on sulphocamphylic acid (CgH,,SO,,H,O = CgH,,O, + H,SO,) is oxidised with permanganate it yields isolauronic acid, CgH,,03, and that the latter acid, when warmed on the water bath with sulphuric acid, is converted into paraxylic acid, C=, (+3 \/ COOH. As these xylic and paraxylic acids may be obtained in this way from sulphocamphylic acid, it seemed possible that the two camphylic acids, C,H120,.might be dihydro-derivative? of these acids and that the acids,AND METHYLTEREPHTHALIC ACIDS, ETC. 159 C9HI4O2, which may be obtained from them by reduction with sodium amalgam, were the corresponding tetrahydro-acids. I n order to deter- mine whether this view was correct, we decided to investigate the action of reducing agents on the xylic acids, and beg t o lay an account, of the results of our experiments before the Society. The xylic and paraxylic acids required for this research were first prepared by the method described by Fittig and Laubinger (Annulen, 1869, 151, 269), which consists in oxidising pseudocumene by means of dilute nitric acid. The product of this reaction contains, besides the two xylic acids, methylisophthalic acid, UH,* C,H,(COOH), [CH, : COOH : COOH = 1 : 2 : 41, rnethylterephthalic acid, and also nitro-derivatives of these acids and of pseudocumene, and as the separation of these various substances by the method given by Fittig and Laubinger is exceedingly tedious, many experiments were made, in the hope of being able to devise a more simple method for preparing these acids. I n the first place, pseudocumene was brominated a t 150' and the bromo-derivative C,H,(CH,),* CH,Br converted into the ethylic ether C,H,(CH,),* CH,* OC,H, by treatment with sodium ethoxide, this was then oxidised in the way recommended by Kipping (Trans., lSS8, 53, 45) for preparing isophthalic acid, but the results were not, satisfactory, and as the irritating vapours of the bromopseudocumene made the process very objectionable, it was abandoned.The next method of preparing paraxylic acid which was investigated was that devised by Armstrong and Kipping (Trans., 1893, 63, 75), who showed that camphor, when digested with sulphuric acid yields acetyl-2-ortho-xylene [Me, : Ac = 1 : 2 : 41, and that this, on oxidation with nitric acid, yields pure paraxylic acid. We, however, soon found that this method is not suitable for the preparation of large quantities of the acid, owing to the difficulty of obtaining acetyl-2-ortho-xylene in a sufficiently pure condition for giving good results on oxidation. Ultimately, we found it necessary to return to the oxidation of pseudocumene with nitric acid, but it very soon became evident that some improvement must be made in the method of separating the products of oxidation, in order to obtain sufficient material for this investigation.The chief objection to Fittig and Laubinger's process is the distilla- tion in steam which they recommend as a means of partially separating the oxidation products, because the xylic acids volatilise with diffi- culty in steam, and therefore, when working with large quantities of160 BEKTLEY AND PERKIN : REDUCTION OF XYLIC, PARAXYLIC, material, this operation takes a considerable length of time and involves the evaporation of large quantities of liquid ; even then, the products obtained are in a very impure condition. Several experiments were made with the object of avoiding this steam distillation, and ultimately the following method was devised, which was found t o work well.After oxidising the pseudocumene in the way recommended by Fittig and Laubinger, the semi-solid product is dissolved in an excess of sodium carbonate, and the nitro-compounds and unaltered pseudocumene removed by extraction with ether. The alkaline solution is then acidified, and the precipitated acids, collected by means of the pump, are dried and con- verted into their methylic salts, which are then fractionated under reduced pressure. I n this way, two principal fractions are obtained, namely, one boiling a t 140-160° (40 mm.), which contains the methylic salts of the two xylic acids, and the other boiling at 180-200" (40 mm.), consist- ing principally of the mixed methylic salts of methylisophthalic acid and methylterephthalic acids ; the former fraction always remained liquid, whilst the latter quickly solidified. The fraction 140-160° (40 mm.) is hydrolysed, the mixed xylic acids thus obtained are converted into their calcium salts, and the latter are then separated by repeated recrystallisation from water ; the separa- tion of methylisophthalic acid and methylterephthalic acid is described later on in this paper. I n our first experiments on the reduction of the xylic acids, the action of sodium amalgam on paraxylic acid was investigated, but even when the alkaline solution of the acid was boiled with sodium amalgam in the way recommended by Aschan (Afinalen, 1892, 271, 234) very little action took place, and nearly all the acid was recovered unchanged.This method was then abandoned, and the reduction of xylic acid and paraxylic acid with sodium and isoamylic alcohol was next tried and found to give satisfactory results, the acids being converted in this way always into a mixture of the tetrahydro- and hexahydro-derivatives ; i t was not found possible to reduce the whole of the acid to the hexa- hydro-derivative, although many experiments were made with this object under very varying conditions, as even after the acids had been subjected repeatedly to the action of the reducing agent, a considerable quantity of tetrahydro-acid was always present. As a rule, after several successive treatments with sodium and isoamylic alcohol, the reduced acid, in each case, was found to contain 50-60 per cent.of the hexahydro- and 40-50 per cent. of the tetrahydro-derivative. On fractionation, the mixed acids distil in each case a t about 160-170O (40 mm.), and the purified product, when leftASD METHYLTEREPHTHALIC ACIDS, ETC. 161 for some time in a freezing mixture, deposits small quantities of solid acid which, on examination, proved to be the tetrahydro-acid. Fetruhyd~~o~u~~ccx~Z~c acid, C,H,,O,, melts a t 83O, and unites with bromine, forming dibromhexahydroparaxylic acid, C9H14Br202, which melts a t 124'. Feti*ahyd~o-xgZic acid, C9HI4O2, melts at I 04O, and also a,bsorbs bromine. The oily reduced acid from both reductions, and from which no more crystals could be obtained, still contains a large proportion of the tetrahydro-cornpound, and attempts were made to convert the whole of it into the hexahydro-acid by repeated reduction with sodium and isoamylic alcohol, and by adding on hydrogen bromide and heating the bromide with zinc dust and acetic acid, but without success.As we were unable to reduce the acids completely to the hexa- hydro-derivatives, we decided to separate the hexahydro-acid from the accompanying tetrahydro-acid, and for this purpose we employed two methods, namely, (1) oxidation of the mixture with potassium perman- ganate, by which means the tehahydro-acids are converted into syrupy hydroxy-compounds, leaving the hexahydro-acids unchanged ; (2) the conversion of the mixed acids into their anilides, by treating the acid chlorides with aniline, and subsequently separating the hexahydro- anilide from the product by fractional crystallisation.The oxidation with permanganate mas performed in very dilute the unaltered hexahydro-acids being subsequently non-volatile hydroxy-acids by distillation in steam. H COOH and cold solutions, separated ifroni the /\ CH, H as a colourless liquid, which on standing solidifies to colonrless crystals melting a t 76-78". H CH, /\ H COOH and does not show any signs of crystallisation, even on long standing. The syrupy hydroxy-acids obtained during the above oxidations could not be obtained in a crystalline condition, and as, on further oxidation VOL. LXXI. M162 BENTLEP AND PERKIN : REDUCTION OF XYLIC, PARAXYLIC, with chromic acid, they did not yield any solid products, their examina- tion was not continued further. It will be seen that, by the above method of separating the hexahydro- acids, the tetrahydro-acids are always destroyed, and as it was especially important t o carefully study these tetrahydro-acids and their derivatives the separation by means of the anilides was tried, but with little success.The acids were converted into the anilides in each case, but on subse- quent crystallisation only the anilides of the hexahydro-acids could be isolated in any quantity in the pure state. This want of success was apparently due t o the presence of a considerable amount of resinous matter, produced, probably, by the action of phosphorus trichloride on the unsaturated tetrahydro-acids ; in both cases, however, besides the anilides of the hexahydro-acids, small amounts of other anilides were isolated, but in quantities too small for further identification.The anilide of hexahydroparaxylic acid melts at 1 1 5 O , whilst that of hexahy- dro-xylic acid melts a t 188'; these anilides are best hydrolysed by heating them with a solution of hydrogen chloride in acetic acid, the yield of acid obtained in this way being quantitative. Prepared by this process, hexahydroparaxylic acid still remained liquid, even when cooled to - 10' ; hexahydro-xylic acid, however, quickly solidified, forming colourless plates which melted at 76-78' : both acids possess a peculiar, pungent smell, somewhat resembling that of the higher fatty acids. Owing t o the difficulty of obtaining the pure tetrahydro-acids in any quantity, experiments were next instituted with the object of preparing these acids from the hexahydro-acids, and in order t o do this the hexahydro-acids were brominated, and the monobromo-derivatives, after conversion into the ethereal salts, hydrolysed, in the first place, with alcoholic potash.It was thought that this method might give very interesting results, as very probably the tetrahydro-acids obtained would be isomeric, and not identical with the tetrahydro-acids prepared by the direct reduction of the xylic acids. However, although the experiments were very much varied, instead of the pure tetrahydro- acids, the products in each case mere semi-solid, uninviting-looking masses, which evidently contained several substances, and as the quantities at our disposal were small, it was impossible t o effect a separation of the various constituents.Apparently, the tetrahydro-acids first formed by this reaction are in part oxidised during the treatment with potash, because, in our experi- ments with hexahydroparaxylic acid, an acid was obtained in this way which had the approximate constitution of a dihydroparaxylic acid. In the case of hexahydro-xylic acid, the bromination did not take place very satisfactorily, as the product, after treatment with alcohol,AXD METHYLTEREPHTHALIC ACIDS, ETC. 163 appeared to contain, besides eth-ylic bromohexahydro-xylate, traces of the ethylic salt of a bromo-xylic acid. The formation of the latter is probably due to the bromine converting part of the hexahydro-acid into xylic acid, which then, by the further action of the bromine, was converted into bromoxylic acid.When the ethereal salts of the bromohexahydro- xylic acids are digested with diethylaniline, they are converted fairly quantitatively into the ethylic salts of the corresponding tetrahydro- acids, which are colourless, sweet-smelling oils, distilling without decomposition at about 1 5 5 O (60 mm.), but these ethereal salts on hydrolysis do not yield the pure tetrahydro-acids ; unfortunately, owing to the nature of the product and to the small quantity of material at our disposal, we were unable to investigate this matter further. Although we have had so much difficulty in obtaining the tetrahy- droderivatives of xylic acid and paraxylic acid in any quantity, we were, nevertheless, able to prepare sufficient of each acid to prove conclusively that these tetrahydro-acids are not identical with the acids which have been obtained by the reduction of the a- and P-camphylic acids.This does not, of course, prove that the a- and P-camphylic acids are not derivatives of the xylic acids, because there are many theoretically possible isomeric tetrahydro-xylic acids which are not described in this paper. As, however, the further study of tho camphylic acids has made it appear unlikely that these acids are closely connected with the xylic acids, we have decided, in the meantime, t o discontinue these experiments, and to publish the results which have so far been obtained. Sepmtion of MetlqZisopAthalic Acid f y o m Nethyltes.ephthalic hi Preparation of Teetralqdrometh ylterephthalic Acid. During the separation of the acids formed by the oxidation of pseudocumene by means of their methylic salts, as explained on page 160, a considerable quantity of a fraction boiling at 180-200' (40 mm.) was obtained, and on investigating this product we were ultimately able to show that it consisted of the methylic salts of methylisophthalic acid, I, and methylterephthalic acid, 11, COOH but no methylphthalic acid could be isolated from the mixture.The above fraction, on cooling, solidified to a crystalline mass, melting roughly a t 40--50°, and by subjecting it to a series of fractional crystallisations, we ultimately succeeded in separating i t into two portions, A, melting at 73-74O, and B, at about 58-60'. Each of these, on analysis, gave M 2164 BENTLEY AND PERKIN : REDUCTION OF XYLIC, PARAXYLIC, numbers agreeing with the formula CH,*C,H,(COOCH,),, and as further recrystallisation did not appear to affect the melting points, we concluded that they were the pure methylic salts of two distinct dibasic acids of the formula CH,* C,H,(COOH),. On hydrolysis with potash, the methylic salt A (m.p. 7 3 - 7 4 O ) yielded an acid melting at 325-330°,* which did not form an anhydride and which, in its properties, agreed so closely with methylisophthalic acid,? that we concluded at first that it was identical with this acid, and that the methylic salt A was methylic methylisophthalate. The methylic salt B (m. p. 58-60'), on hydrolysis, yielded an acid, CH,*C,H,(COOH),, which melted a t 293-295' and gave no anhydride, and consequently we supposed it to be .IrLetl7LyZte~~eio~t~~a~~c cmk?, especially as Fittig and Laubinger (Annalen, 1869, 151, 276) had described methylterephthalic acid as melting at 280-283", and, after sublimation, at 291'. On treating the supposed methylisophthalic and methyltereph thalic acids with sodium amalgam, it was observed that the acid (m.p. 325- 330") corresponding with the methylic salt A (m. p. 73-74°) was readily and completely reduced to a tetrahydro-acid, whereas the acid (m. p. 293-295') obtained from the methylic salt B (m. p. 58-60°) was only partially reduced, yielding a tetrahydro-acid which was evidently identical with that obtained from the acid of higher melting point. The acid which had not been reduced during this treatment was carefully purified, and found to have the formula CH,* C,H,(COOH), ; it melted now at 320-33,0°, and yielded a methylic salt which fused sharply a t SO', and which may be called C.This acid is not in the least degree affected by boiling with sodium amalgam, and, as it does not give an anhydride, it must obviously be either methylisophthalic acid or methylterephthalic acid, and the acid which Fittig and Laubinger described as methylterephthalic acid must be a mixture of this acid with methylisophthalic acid, one of which is reduced by sodium amalgam, leaving the other entirely unattacked. I n order to determine which of these two acids is reducible by sodium amalgam, we made a number of experiments, with the object of IC With regard to the melting point of methylisophthalic acid, compare Clam, J.gr, Chem., 1890, [ 2 ] , 42, 510. j. Methylisophthalic acid has been prepared by Jacobsen (Ew., 1881, 14, 2112) by the oxidation 1 : 4-dimethylbenzoic acid with permangsnate ; by Claus (Ber., 1886, 19, 233 ; J. pr. Chem., 1890, [ 2 ] , 42, 509) by the oxidation of methyl cymyl ketone, CH,*CO.C,H,(CH,)'C,H,, and of the acid CH,*C,H,(C,H,)~CO~COOH [CH,: CO : C,H,= 1 : 3 : 51, and by Hjelt and Gadd (Ber., 1886, 19, 868) by the oxidation of pscudocumenyl alcohol, CH3.C6H3(CH,*OH),.AND METHYLTEREPHTHALIC ACIDS, ETC. 165 deciding which of the two acids formed a methylic salt melting a t 73-74", and which a mcthylic sa,lt melting a t 80'. By oxidising xylic acid and paraxylic acid with potassium perman- ganate, we obtained, in both cases, an acid melting a t 325-330" which did not give an anhydride, and which must, therefore, be metlqyZtes*eph- thulic acid.coo H CH, COOH f \ C H 3 and f j c ~ 3 give \j \/ COOH \/ COOH CH3 As the methylic salt of this acid melts a t 73-74', it follows that the methylic salt A is methyZic terephthdute. When iso-xylic acid is oxidised with permanganate, it yields a dibasic acid which does not form an anhydride, and which must, therefore, be meth$isopI~thalic acid. CH3 CH3 OCooH- ( Y O O H gives COOH \/ CH, This acid melts a t 320-330', and yields a methylic salt melting at SO". On treatment with sodium amalgam at 1 O O O , met'hylisophthalic acid remains unchanged, whereas, on the other hand, mebhylterephthalic acid, under the same conditions, is apparently completely reduced to a tetrahydro-derivative.T~trahydrornethylterep~~t~alic acid, CH,. C,H7(COOH),, is moderately easily soluble in water, and melts a t 240-245'; it yields a dimethylic salt, CH,* C,H7(COOCH,),, which is a colourless oil boiling a t 165-170' (20 mm.). Although the solution of the acid in sodium carbonate very quickly decolourisea permanganate, pet the free acid does not appear to absorb bromine or hydrogen bromide a t all readily. When heated with a saturated solution of hydrogen bromide in glacial acetic acid a t 100' in a sealed tube, it yields a black, oily maPs containing bromine, but which could not be purified. Experiments on the reduction of this crude hydrobromide with zinc dust and acetic acid were instituted, but we have so far not been able to isolate the hexahydromethyltere- phthalic acid, which should be formed in this way.The methylic salt C is, theref ore, meth?jlic isophthabte. EXPERIMENTAL. Oxidution of Pseudocumene.-For this purpose, as stated in the introduction, we employed a modification of the process devised by166 BENTLEY AND PERKIN : REDUCTION OF XYLIC, PARAXYLIC, Fittig and Laubinger (Ann., 1869, 151, 269), their method of separat- ing the products by distilling in a current of steam being very trouble- some and inconvenient when working with large quantities. As the result of numerous experiments, we ultimately adopted the following method in preparing the acids required for this research. Pseudocumene (250 grams), mixed with dilute nitric acid (700 grams of acid, sp. gr. 3.4, diluted with 24 volumes of water), is heated to boiling for 18 hours in a large flask connected with a reflux condenser.The cold liquid is then largely diluted, the semi-eoli d precipitate collected with the aid of the pump, washed well with water, and dissolved in a solution of sodium carbonate. The nitro-compounds and un- altered pseudocumene which remain undissolved are removed by agitation with ether, and the aqueous solution is heated to boiling and acidified with hydrochloric acid. The precipitate thus obtained is col- lected, washed with water, and after being pressed on porous plates to remove oily matter, is dried a t 100'. The mixed acids are then con- verted into their methylic salts by saturating their solution in methylic alcohol with dry hydrogen chloride ; the mixed methylic salts, after being precipitated with water and extracted with ether, &c., in the usual way, are fractionated several times under a pressure of 40 mm.Of the two fractions obtained, the larger distilling at 140-155', consists princi- pally of the methylic salts of xylic and paraxylic acids; whilst the smaller fraction, which distils at 1 80-200°, and quickly solidifies on cooling, contains the methylic salts of methylterephthalic and methylisophthalic acids. A third, but very small, fraction, boiling at 250-270', was also obtained; this was a very viscid liquid, which possibly contained the methylic salts of trimellitic acid, but it was not further investigated. Xepwation of the Xylic and Pamxylic Acids.--In order t o isolate these acids, the fraction (135 grams) containing their methylic salts is hydrolysed by boiling with a solution of alcoholic potash (80 grams) in a reflux apparatus for 3 hours, the alcohol distilled off, the residue dissolved in water, and the solution evaporated to a moderately small bulk.While still hot, t.his is acidified with hydrochloric acid, and the precipitate, which is somewhat oily a t first, but on cooling becomes quite hard, is first roughly freed from oily matter and other impurities by crystallisation from acetic acid; if his is not done, the subsequent separation of the isomerides by means of their calcium salts becomes a very difficult matter. The mixture of acids which separates from the hot, concentrated solution in glacial acetic acid in the form of a crystalline precipitate melting roughly at 1 0 5 O , is collected, washed first with acetic acid and then with water, and dis- solved in a slight excess of hot sodium carbonate solution; on mixingAND METHYLTEREPHTHALIC ACIDS, ETC.167 this with a strong, hot solution of rather more than the calculated quantity of calcium chloride, and cooling, a mixture of the calcium salts of the two xylic acids separates almost completely; the salts, after being collected and washed with a little water, are crystallised from a considerable quantity of boiling water. If the acids are puri- fied as described above, the f i s t crystallisation yields the almost pure calcium salt of paraxylic acid, this being the least soluble. It is, however, a matter of considerable difficulty to separate the calcium salts contained in the mother liquor, but this may be effected in the following manner.When the solutions of the calcium salts have yielded as much pure calcium paraxylate as possible, the mother liquors are acidified, and the precipitated acids collected and crystallised from acetic acid. A considerable quantity of nearly pure xylic acid separates first in beautiful, long needles. After this has been collected, and the acetic acid removed from the mother liquors by distillation, the acids are again converted into the calcium salts and crystallised from water. By a repetition of this process, it is possible to separate the whole of the material into calcium paraxylate and xylic acid. The paraxylic acid obtained from the pure calcium salts by precipitation with hydrochloric acid was recrystallised from acetic aci-d.Xylic acid, U,H,(CH,),* COOH [CH, : CH, : COOH = 1 : 3 : 41, crystallises from acetic acid in beautiful, long, prismatic needles melting at 126'. Paraxylic acid, [CH, : OH, : COOH = 1 : 2 : 41, crystnllises in short needles melting a t 163'. As stated in the introduction, this acid is one of the products of the reduction of paraxylic acid by means of sodium and isoamylic alcohol. Pure paraxylic acid (30 grams) is dissolved in isoamylic alcohol (1% litres) contained in a large, round-bottomed flask attached to a long reflux condenser. The liquid is heated to the boiling point, and then sodium (10 grams) is dropped in ; a violent action takes place, and as soon as this has subsided more sodium (20.grams) is added. When the sodium has all dissolved (the mixture being heated if necessary), the contents of the flask are allowed to cool, diluted with water and acidifled with hydrochloric acid; the liquid then separates into two layers, the upper of which is isoamylic alcohol containing the organic acid in solution. The lower layer, consisting of sodium chloride soln- tion and some hydrochloric acid, is shaken with pure isoamylic alcohol, in order to extract any traces of the organic acid which it may contain, and the whole of the isoamylic alcohol solution is dehydrated by dis- tillation until the temperature of the vapour rises to 120'. It is then treated with sodium (30 grams) exactly as before, and the process of168 BENTLEY AND PEHKIN : REDUCTION OF XTLIC, PARAXTLIC, extraction again repeated.After thrice treating with sodium, the product is diluted with a large quantity of water, the upper layer of isoamylic alcohol is distilled almost to dryness, and the residue dissolved in water and mixed with the aqueous layer. I f this is not done, there is a very considerable loss of product, as the isoamylic alcohol retains a quantity of the sodium salts of the reduced acids, even after it has been repeatedly washed with water, The aqueous solution, after being boiled until free from isoamylic alcohol, is acidified and extracted with pure ether, the ethereal solution is dried over calcium chloride, the ether distilled off, and the oily residue fractionated under diminished pressure. Nearly the whole distils at about 160-170° under a pressure of 40 mm.When, however, about three-quarters of the acid has distilled, the receiver is changed and the last portion collected separately ; the latter, when exposed to the cold for about a week, deposits crystals of tetrahydro- paraxylic acid, whereas the first fraction if similarly treated yields very few crystals. The crystals were collected with the aid of the pump, pressed on a porous plate, and recrystallised several times from light petroleum (b. p. 60-80"). I. 0.1020 gave 0.2618 CO, and 0.0830 H,O. C = 70.00 ; H = 9.04. 11. 0.1096 ,, 0.2820 CO, ,, 0*0900 H,O. C = 70.17 ; H = 9.12. C6H7(CH,),.COOH requires C = 70.13 ; H = 9.09 per cent. It is easily soluble in benzene, alcohol, ether, and acetic acid ; sparingly so in cold, easily in hot, light petroleum.I t s solution in sodium carbonate immediately reduces potassium permanganate in the cold, and bromine is quickly absorbed by a solution of the acid in chloroform. The yield of reduced acid, after distillation, was usually about 75 per cent. of the theoretical quantity, but of this only about 10 per cent. separated in the form of crystals of tetrahydroparaxylic acid. Tetruhydroparuxylic cccid crystallises in prisms melting a t 83". Dibromhexcc?~yds.o~aruxylic Acid, CSH1,Br,O,. This was prepared by dissolving pure tetrahydroparaxylic acid in dry chloroform and adding a chloroform solution of bromine until the colour of the bromine remained permanent. The oil which was deposited, on allowing the chloroform to evaporate spontaneously, slowly solidified.This crude product purified by recrystallisation from dilute acetic acid saturated with hydrogen bromide, separated in colourless plates melting at 124' 0.0532 gave 0.0640 AgBr. Br=51*17. Dibromhexahydro~uraxyEic cccid melts a t 124". C9H,,Br,02 requires Br = 50.95 per cent. It dissolves readily in most organic solvents on boiling, being apparently decomposed at the same time, as we could not succeed in recovering it from them inAND METHYLTEREPHTHALIC ACIDS, ETC. 169 a crystalline form. It separates, however, readily from dilute acetic acid if decomposition is prevented by previously saturating the solution with hydrogen bromide. NexcchydropcwaxyZic acid, C,H,(CH,),*COOH, CH, I CH. CH, PH\\ Hzf! I II,C CH, ‘CH” I COOH This acid is formed when paraxylic acid is reduced by sodium and iso- amylic alcohol as explained on page 167, and is present in considerable quantities in the mother liquor of the reduced acid after the bulk of the crystalline tetrahydroparaxylic acid has been removed by cooling in a freezing mixture.I n order to isolate it, two methods were employed. 1. The crude product was treated with permanganate so as to oxidise the tetrahydroparaxylic acid, the hexahydroparaxylic acid remaining unchanged. 2. The mixed acids were converted into the anilides, and these sepa- rated by fractional crystallisation. iMetftod I.-The oxidation mas carried out in the following manner. The oily mixture of tetrahydro- and hexahydro-paraxylic acids was dis- solved in a small quantity of sodium carbonate solution, diluted with a large quantity of water, and oxidised by running in slowly a weak solution of potassium permanganate.During the whole operation, a rapid current of carbon dioxide was passed through the liquid, which was vigorously stirred by means of a turbine, and kept cold by adding ice from time to time. When the colour of the permanganate re- mained permanent, the liquid was mixed with a little alcohol, boiled, and the manganese dioxide filtered off, well washed with hot water, and the filtrate evaporated to a small bulk. This was then acidified and subjected to distillation in a rapid current of steam until the condensed water was quite clear and had only a very slight acid reaction. The dis- tillate was extracted three times with ether, and the oily residue left on evaporating the dried ethereal solution was fractionated at the ordi- nary pressure. Nearly the whole distilled at 251’ (748 mm.) as a colourless oil which gave the following results on analysis.0.1480 gave 0.3760 CO, and 0.1336 H,O. C = 69.28 ; H = 10.03. CGH9(CH,),* COOH rcquires C = 69.23 ; H = 10.25 per cent.170 BENTLEY AND PERKIN : REDUCTION OF XYLIC, PARAXYLIC, Hexahydroparaxylic acid is an oil of peculiar odour, somewhat similar to that of the higher fatty acids; it remains liquid even when cooled to a temperature of - 10". Method II.-The separation of hexahydroparaxylic acid by means of its anilide was conducted as follows. The crude mixture of acids (30 grams) was first converted into the acid chlorides by heating with phos- phorus trichloride (15 grams) in a reflux apparatus for 10 minutes ; the liquid was then decanted from the layer of phosphorus acid, and distilled under reduced pressure.The fraction boiling a t 135-145' (60 mm.) was dissolved in pure, dry ether, and slowly mixed with an ethe- real solution of aniline (60 grams) ; after standing some time, water was added, the ethereal solution separated, washed with dilute hydro- chloric acid until free from aniline, dried, and the ether distilled off. The oily residue, on standing, quickly solidified, and the product, after two crystallisations from light petroleum (b. p. 100--120°), was ob- tained in colourless prisms melting at 115'. This substance is the anilide of hexahydroparaxylic acid, as was proved by comparing it with a sample of the anilide prepared from the pure acid.I n order to obtain the hexahydro-acid from its anilicle, the latter is heated for 12 hours a t 150' in a sealed tube with acetic acid saturated with hydrogen chloride, and containing a little aqueous hydrochloric acid. The pro- duct was then diluted with water and extracted several times with ether. I n order to remove the aniline, the ethereal solution was shaken with excess of sodium carbonate solution, and the alkaline liquid acidified and again extracted with ether. This ethereal solution was washed, dried, and evaporated, the oily residue being purified by fractional distillation. Even when prepared from the pure anilide in this way, hexahydroparaxylic acid is an oil which does not solidify, even at low temperatures.The yield of hexahydroparaxylic acid obtained from the crude mixture of acids by this method is from 30 to 40 per cent. The petroleum mother liquors of the anilide of hexahydroparaxylic acid, after the removal of the light petroleum by distillation, deposited an oil from which a small quantity of solid separated on standing. This, after being crystallised from petroleum, melted a t 140-145" ; the quantity, however, was too small for further examination. The attempts made with the object of obtaining tetrahydroparaxylic acid from the oily anilide which remained after the above crystalline products had been extracted as completely as possible, were unsatis- factory. Hexahy dr opcwaxy Zp? ch Zoyide, C,H,( CH,), COO1. This was prepared by gently heating pure hexahydroparaxylic acid (4 grams) with phosphorus trichloride (2 grams) for 10 minutes ; the liquid was decanted from the phosphorous acid and distilled under reduced pressure.It is a disagreeably-smelling liquid boiling at 110" (25 mm.).AND METHYLTEHEPHTHALIC ACIDS, ETC. 171 0.2406 gave 0.1996 Ag C1. C1=20*52. C8H,,*COC1 requires C1= 20.34 per cent. Eth ylic hexuh ydropccrccx ylcte, C,H, (CH,),. COOC,H5. This was prepared by pouring the acid chloride into three times its volume of alcohol, and after allowing it to stand a short time, diluting with water, extracting with ether, &c. The brownish oil thus obtained was purified by fractional distillation. Ethylic hexahydroparaxylate is an oil of pleasant odour, and lighter than water; it boils at 224' (758 mm.).0.12'72 gave 0.3334 CO, and 0.1210 H,O. C = 71.48 ; H = 10.65. C,H,,*COOC,H, requires C = 71-74 ; H = 10.87 per cent. Artilide of hexuhydyopuraxyZic w i d , C,H,,* CO* NH- C6H5. This anilide was prepared by mixing ethereal solutions of equal quantities of the acid chloride and aniline. After standing for some time, water was added, and the ethereal solution washed first several times with small quantities of dilute hydrochloric acid until free from aniline, and then once with water. It was then dried andthe ether distilled off; the oily residue gradually solidified, and after being purified by crystallisation from light petroleum (b. p. SO--looo) the. anilide was obtained in prisms melting a t 115'. N = 6-39 ; C,H,,* CO *NH* C,H, requires N = 6-06.The anilide of hexahydroparaxylic acid is moderately soluble in ben-- zene and alcohol, but only slightly in cold light petroleum (b. p. 80- loo'), although it dissolves readily in the boiling solvent. Analysis ; found Etlqlic bronahexahyd~.opi~raxylute, C,HsBr(CH,)2* COOEt YH3 CH* CH, P"\ I I COOC,H, This was prepared, for reasons explained in the introduction, by brominating hexahydroparaxylic acid and subsequently pouring the product into absolute alcohol. Hexahydroparaxylic acid (5 grams) was carefully mixed with phosphorus pentabromide (5 grams) and bromine (6 grams), and gently heated on the water bath until the bromine had disappeared. The product, poured into twice its volume of alcohol, was allowed to stand for some time, and then diluted with water, the heavy oil which172 BENTJ,EY AND PERKIN : REDUCTION OF XYLIC, PARAXYLIC, separated being extracted with ether, kc., in the usual manner.The residue thus obtained, when fractionated under reduced pressure, yielded ethylic bromhexahydroparaxylate as a heavy oil boiling without decompo- sition a t 170-180' (55 mm.); the yield was about 70-80 per cent. 0.1928 gave 0.1376 AgBr. C,H,,Br* COOC,H, requires Br = 30.112 per cent. When hydrolysed with potash, this ethereal salt yields a mixture of acids, a behaviour which is probably due in part to the oxidation of the tetrahydroparaxylic acid formed in the first instance. The only sub- stance we were able to isolate from this mixture was a small quantity of a crystalline acid which melted roughly a t 135-140°, and on analysis gave numbers corresponding with those required by dihydro- paraxylic acid, C,H,,O,.Br = 30.37. Ethylic tetrahydroparaxylute, C,H7( CH,),* COOC,H,. This is formed when ethylic bromhexahydroparaxylate is digested with diethylaniline. Pure ethylic bromhexahydroparaxylate mixed with twice its weight of diethylaniline was heated to gentle ebullition €or 4 hours, and then poured into dilute hydrochloric acid. The dark- coloured oil which separated was extracted with ether, washed, and treated in the usual way. The product on being fractionated under re- duced pressure, yielded a colourless, sweet-smelling oil which distilled at 155' (60' mm.), and on analysis gave numbers showing that it was of ethylic tetrahydroparaxylate. 0-0814 gave 0.2162 CO, and 0.0720 H,O.C,H,,*COOC,H, requires C = 72.52 ; H = 9.89 per cent. I n order, if possible, to prepare the corresponding tetrahydroparaxylic acid, the ethylic salt was hydrolysed. Ethylic tetrahydroparaxylate (3 grams) was heated on the water bath with potash (3 grams) dissolved in methylic alcohol for 2 hours, the alcohol distilled off, the residue evaporated with water until free from alcohol, and then acidified and extracted with ether. The dried ethereal solution, on distillation, left an oil which partially solidified. The crystals, after being pressed upon a porous plate and crystallised from light petroleum (b. p. 100-120°), formed prisms melting a t 135-140°, and on analysis gave numbers agreeing with the formula C,H,,O,. C = 71.14; H = 7.93. C = 72.43 ; H= 9.82.0.1180 gram gave 0.3068 GO, and 0.0842 H,O. CgH,,O2 requires C = 71.05 ; H = 7.88. C,H,,O, requires C = 70.13 ; H = 9.09 per cent. This substance is, therefore, probably a dihydroparaxylic acid identical with the substance described above.AND METHYLTEREPETHALIC ACIDS, ETC. 173 Reduction of Xylic Acid. The reduction of xylic acid with sodivm and isoamylic alcohol, and the methods used in the separation of t,he acids thus formed, were con- ducted in almost exactly the same manner as in the case of paraxylic acid ; a brief description of the process, therefore, is all that is needful. Pure xylic acid (30 grams), dissolved in isoamylic alcohol (I$ litres), was heated t o boiling and treated three times in succession with sodium (30 grams), the same precautions and methods of extraction being adopted as in the reduction of paraxylic acid.The oily acid thus obtained when fractionated under reduced pressure passed over entirely a t about 160--170° (40 mni.). When three-quarters of the acid had distilled, the last portion was collected separately ; this soon deposited crystals when exposed to the cold, whilst the first fract.ion usually remained liquid. Tetvahydyo-xylic Acid, C,H7(CH3),*COOH. This was obtained from the higher fraction of the reduced xylic acid, the crystals which were deposited being pressed, and recrystallised several times from light petroleum (b. p. SO-looo). It forms plates which melt a t 103’. 0.1000 gave 0.2570 CO, and 0.0830 H,O. C,H7(CH3),*COOH requires C = 70.13; H = 9.09 per cent. Tetrahydro-xylic acid is readily soluble in benzene, alcohol, and acetic acid; sparingly so in cold petroleum, but very soluble in the hot solvent.When dissolved in sodium carbonate solution, it instantly decolorises potassium permanganate ; its solution in chloroform also decolorises bromine very readily, but, unfortunately, we were not able to isolate the product owing to the small amount of material a t our disposal. He.xcchyZi*o-xyZic Acid, C,H,(CH,),*COOH. C = 70.09 ; H = 9.22. COOH I PH\ CH*CH, I CH, This acid, like the corresponding hexahydroparaxylic acid, was isolated from the mixture of reduced acids by two processes, viz., (1) oxidation with potassium permanganate, and ( 2 ) hydrolysis of the anilide. I n both cases, a liquid is obtained which, when prepared from the anilide, solidifies quickly, whereas the product obtained by oxidation solidifies but slowly and vcry incompletely. The second method gives the purer product.174 BENTLEY AND PERKIN : REDUCTION OF XYLIC, PARAXYLIC, 0.1694 gave 0.4288 CO, and 0.1568 H,O.Hexahydro-xylic acid crystallises from light petroleum (I 00-1 20') in thick plates melting a t 76-78". It boils at 250-255", and is readily soluble in benzene, alcohol, and acetic acid ; sparingly in cold, easily in hot, light petroleum. I t s solution in cold dilute sodium carbonate decolorises permanganate only very slowly. C = 69.03 ; H = 10.28. C,H9(CH,),COOH requires C = 69.23; H = 10.25 per cent. Anilide of Hexahy&o-xylic Acid, C,H9(CH,),*CO*NH*C,H,. This anilide was obtained from reduced xylic acid, and employed for the preparation of the foregoing hexahydro-xylic acid.Reduced xylic acid (28 grams) was heated for about 15 minutes with phosphorus trichloride (14 grams), and the liquid decanted from the phosphorous acid and distilled under reduced pressure ; the acid chloride (30 grams), which came over a t about 130-140" (40 mm.), was then dissolved in pure, dry ether, and carefully added to an ethereal solution of aniline (56 grams); after a short time water was added and the ethereal solution washed, dried, &c., as before. The oil obtained in this way soon solidified, and after being four times recrystallised from a mixture of alcohol and petroleum (b.p. 100-120"), was obtained in needles melting a t 180'. On analysis, nitrogen was found = 6.26 per cent.; C,H,,- CO *NH* C,H, requires N r= 6-06 per cent.The anilide of hexah?/dyo-xylic acid is sparingly soluble in cold ben- zene and alcohol, but dissolves readily in these solvents on boiling ; i t is almost insoluble in cold petroleum (b.p. 100-120°) and only sparingly soluble in the hot solvent. These properties, together with its unusually high melting point, distinguish this anilide from the cor- responding anilide of hexahydropnraxylic acid. It is hydrolysed on heating with an acetic acid solution of hydrogen chloride in a sealed tube for 12 hours at 150", an almost quantitative yield of hexahydro-xylic acid being produced. Ethylic 61wom~exa~~yd.1.o-xyZute, C,H,,Br*COOC,H5. COOC,H, I CH* CH, PBr\ H27 H,C CH2 CH3 'CR' I This mas prepared in the same manner as the isomeric ethylic brom- hexahydroparaxylate.The specimen prepared boiled a t 160-1 70"AND OF XETHYLTEREPHTHALIC ACIDS, ETC. 175 (40 mm.), but was apparently not quite pure and probably contained traces of the ethylic salt of bromo-xylic acid, C,H,Br*COOC,H,. We were led to these conclusions from the fact that a bromine determina- tion gave too high a result, and, secondly, that the oil on hydrolysis with potash yielded a small quantity of a solid acid which melted at 170" and contained bromine, apparently a monobrorno-xylic acid. The ethylic bromhexahydro-xylate, prepared as explained above, gave the following results on analysis. 0.3010 gave 0.2222 AgBr. Br = 31-41. CsH,4Br*COOC,H5 requires Br = 30.42 per cent. Ethylic tetmhydyo-xybte, C,H,,* COOC,H,.This was obtained by the action of diethylaniline on the brom- It is a pleasant-smelling liquid boiling ethereal salt just described. a t 228O (752 mm.). 0.1293 gave 0.3418 CO, and 0.1122 H,O. C,H,,* COOC,H, requires C = 72.52 ; H = 9.89 per cent. A small quantity of this ethylic salt was hydrolysed with potash and yielded an acid melting at 98-110", but the quantity was insufficient for further examination. C =72*02 H= 9.64. Pt*epcwcction of Methylisophthalic Acid und of Methylterepltthcclc Acid. As was explained in the introduction, the oxidation of pseudocumene with dilute nitric acid gives rise to two dibasic acids, namely, methyl- isophthalic acid, CH,. C,H,(COOH), [CH, : COOH : COOH = 1 : 2 : 41, and methylterephthalic acid, and when the acids contained in the crude product of the oxidation are converted into their methylic salts and these are fractionated (see p.160), the fraction lS0-200' (40 mm.), consists almost entirely of the methylic salts of these two acids. This fraction solidifies on cooling, and the solid mass, when subjected to a series of crystallisations from methylic alcohol, separates into two portions, A, the least soluble, melting a t 73-74', and B, which melts at about 58-60'. The melting point of the fraction B does not rise on further crystallisation, and it was, therefore, at first supposed that this substance was pure, especially as, on analysis, it gave numbers agree- ing closely with those required by the formula CH,* C,H,(COOCH,),. 0.1280 gave 0.2986 GO, and 0.0686 H,O. CH,*C,H,(COOCH,), requires C = 63.46 ; H = 5.76 per cent.Further experiments showed, however, that it is a mixture of methylic methylisophthalate and methylic methylterephthalate, and the only way which we could find to separate the methylisophthalic from the methylterephthalic acid was by treatment with sodium amalgam. C = 63.62 ; H = 5.95.176 BENTLEY AND PERKIN : REDUCTION OF XYLTC, PARAXTLIC, Met?~yliso~~?~tl~clic Acid, [CH, : (COOH), = 1 : 2 : 41. I n order to prepare this acid, the methylic salts B, melting a t 58-60' (70 grams) were dissolved in alcohol, mixed with an alcoholic solution of potash (65 grams), and heated on the water bath for 2 hours. The alcohol was then distilled off, the residue diluted with water, evaporated until entirely free from alcohol, and the aqueous solution, after filtration, was acidified ; the copious, flocculent precipitate which separated was collocted with tho aid of the pump, washed well with water, and dried, first on porous plates, and then a t 100".It melted a t 285-290', but on repeated recrystallisation from glacial acetic acid the melting point rose to 293-295", but no higher. This substance is a mixture of methylisophthalic and methylterephthalic acids, and in order to obtain the former acid from it in a pure state, the mixture, in quantities of 10 grams at a time, was dissolved in sodium carbonate solution in a strong porcelain beaker and 3 per cent. sodium amalgam (1 kilo.) added in quantities of 100 grams a t a time. The temperature of the reaction was kept at about 100" by immersing the beaker in a boiling water bath, a rapid current of carbon dioxide was also passed into the mixture during the whole operation, and the separation of sodium salts was prevented by the addition from time to time of small quantities of boiling water.When the sodium amalgam had all been used up, the aqueous solu- tion was decanted from the mercury, filtered, and acidified; the bulky precipitate then thrown down consisted of nearly pure methylisophthalic acid, as the tetrahydromethylterephthalic acid which is formed during the reduction is soluble in water, and remains dissolved in the mother liquor. The precipitate was collected with aid of the pump, waslied well with water, dried first on a porous plate, and then at looo, and purified by conversion into the methylic salt.For this purpose, the well-powdered dry acid was suspended in methylic alcohol, the mixture saturated with hydrogen chloride and allowed to stand for some hours; water was then added and the methylic salt extracted with ether. The ethereal solution, after being washed with water and sodium carbonate solution, was dried over calcium chloride, and the ether distilled off, when an oil was left which quickly solidified. After crystallising three times from methylic alcohol, the methylic salt was obtained pure in the form of prismatic needles, which, when heated in a capillary tube, sintered somewhat a t 77", and melted a t 80'. I. 0-1592 gave 0.3710 CO, and 0.0864 H,O. C=63*55; H=6*03. II. 0.1172 ,, 0.2726 CO, ,, 0.0628 H,O. C=63*43; H=5*95.CH,* C,H,(COOCH,), requires C = 63-46 ; H = 5.76 per cent. Methylic methylisophthalate is readily soluble in most of the ordinary organic solvents, it is, however, but sparingly soluble in cold methylicAND O F METHYLTEREPHTHALIC ACIDS, ETC. 177 or ethylic alcohol, although it dissolves readily in these solvents on boiling. On hydrolysis with potash, it yields pure ~ ~ t ~ y Z i s o p ~ ~ t ~ u Z ~ acid, which separates from acetic acid as a white, apparently amorphous, powder, and melts at about 320-330". 0.1188 gave 0.2618 CO, and 0.0480 H,O. CH;C,H,(COOH), requires C = 60.00 ; H = 4.44 per cent. Methylisophthalic acid is practically insoluble in most organic sol- vents ; it dissolves slightly, however, in boiling water, and, on cooling, separates almost completely in white, flocculent masses.When treated with acetic anhydride or acetyl chloride, it does not yield an anhydride. Prolonged treatment with sodium amalgam does not appear to have any action on this acid, nearly the whole being recovered unchanged on acidifying the alkaline product of the reaction. I n purifying methylisophthalic acid in the way described above, the tetrahydroterephthalic acid formed during the reduction remains dissolved in the liquors from which the former acid separates on acidi- fication. To obtain this tetrahydro-acid, the solution was evaporated to dryness, and the residue extracted with ether in a Soxhlet apparatus ; after distilling off the ether and crystallising the extract from water, pure tetrahydromethylterephthalic acid was obt-ained, melting at 240-245" (see p.178). C = 60.01 ; H = 4.48. 0.11 10 gave 0.2386 CO, and 0,0642 H,O. C = 58.62 ; H = 6-42. CH,* C,H;(COOH), requires C = 58.69 ; H = 6.62 per cent. i ~ ~ ~ t l ~ ~ Z t e l . e y , l ~ t h c ~ Z ~ c Acid, [Me : (COOH), = 1 : 2 : 51 The methylic salt A, the preparation of which is given on p. 163, consists of pure methylic met h ylt erephthalate. 0.1052 gave 0.2450 CO, and 0.0540 H,O. C=63-51; H=5.70. CH,-C,H,(COOCH,), requires C = 63.46 ; H = 5.76 per cent. Methylic ~,zethyltes.e~htl~alcl.te is very readily soluble in benzene, ethylic acetate, light petroleum, and hot alcohol, but comparatively sparingly in cold methylic or ethylic alcohol. In order to prepare nzethylterephthalic acid, the pure methylic salt (7 grams) was dissolved in alcohol, and heated on the water bath with alcoholic potash (7 grams) for 2 hours.The product was diluted with water, evaporated until free from alcohol, filtered, and acidified, when a very voluminous precipitate was thrown down. This was collected with the aid of the pump, washed well with water, dried, and purified by crystallisation from hot glacial acetic acid, from which it separates as an apparently amorphous powder melting at 325-330'. It melts at 73-76'. VOL. LXXI. N178 BENTLEY AJSD PERKIK : HEDUCTION O F XYLIC, PARAXYLIC, 0.1106 gave 0.2420 CO, and 0.0458 H,O. I~eth?/Zterep~~thacEic acid resembles methylisophthalic acid very closely in its properties. It is practically insoluble in most organic solvents, such as benzene, chloroform, light petroleum, and ether, and although more readily soluble in boiling xylene and glacial acetic acid, it is practi- cally insoluble in these solvents in the cold.When heated, it sublimes without forming an anhydride. Methylterephthalic acid is moderately easily reduced when its solution in sodium carbonate is boiled with sodium amalgam, and in this respect it differs remarkably from methylisophthalic acid, which does not appear to be affected a t all by this treatment. C=59*67; H=4-61. CH,* C,H,(COOH), requires C = 60.00 ; H : 4-44. per cent. ~etrcc~?/drometl~yhterepl~~~aEic Acid, CH,*C,H7(COOH),. I n order to prepare this acid, methylterephthalic acid (18 grams) was dissolved in a little sodium carbonate solution, heated to about 100" by means of a water bath, and sodium amalgam (2 kilos.) added in small quantities at a time, a current of carbon dioxide being passed through the liquid during the operation. When the sodium amalgam had all been added, the aqueous solution was filtered, acidified, and the precipi- tate, collected by the aid of the filter-pump, was washed with a little water, and purified by crystallisation from hot water. It is thus obtained apparently in the amorphous condition ; it melts at 240-245", sintering a t about 230".0.1 182 gave 0.2554 CO, and 0.0696 H,O. CH,*C,H, (COOH), requires C = 58.69 ; H = 6.52 per cent. ~etra~~ydiwonzethylterep~~t~~alic acid is fairly soluble in organic solvents, and especially readily in glacial acetic acid ; it is easily soluble i n hot water,-and as it is also fairly soluble in cold water, a considerable quan- tity remains in the mother liquors after the acid has been collected.This can be recovered by evaporating and extracting the solid residue with ether in a Soxhlet's apparatus (see p. 177). A solution of the acid in sodium carbonate decolourises potassium permanganate very quickly, but the acid itself appears t o have little affinity for bromine or hydrogen bromide. An attempt was made to prepare the bromhexahydro-acid by heating the tetrahydro-acid with a solution of hydrogen bromide in glacial acetic acid in a sealed tube a t 100" for several hours. On diluting the product with water, a black, oily-looking precipitate was obtained, but all attempts to isolate a pure bromhexahydro-acid for it were unsuc- cessful. When the black mass was dissolved in acetic acid and treated C = 58.82 ; H = 6.53.AND OF METHYLTEREPHTHALIC ACIDS, ETC.179 with zinc dust, a small quantity of what appeared to be impure tetra- hydromethylterephthalic acid was obtained. The experiment was repeated several times, and in one instance an acid was obtained which decomposed permanganate only very slowly, and melted a t 210-220" ; the quantity, however, was too small t o permit of further investigation. This compound mas prepared by saturating a solution of the acid in methylic alcohol with dry hydrogen chloride, and adding water after the mixture had been left for several hours. The oil which separated was extracted with ether, the ethereal solution washed with sodium car- bonate and then with water, and dried by calcium chloride ; the ether was distilled off and the oily residue distilled under reduced pressure.C = 61.83 ; H = 7-59, 0.1676 gave 0.3800 CO, and 0.1146 H,O. Methylic tet.r.c~hydromethyEt~~ep~~tl~abte is an oil possessing compara- CH,*C,H7(COOCH,), requires C = 62.26; H = 7.55 per cent. tively little odour ; it boils at 165-170' (20 mm.). Preprution of i~et~&yltei.e~~~t~~uZic A cid 69 the Oxidation of Xylic Acid und of Paraxglic Acid. These experiments were instituted, as explained in the introduction, with the object of obtaining pure methylterephthalic acid for comparison with the acids obtained by the oxidation of pseudocumene. Xylic acid (2 grams) was dissolved in sodium carbonate solution and heated on a water bath with a dilute solution of potassium permanganate (4 grams) until the pink colour had disappeared.The manganese dioxide was then removed by filtration, washed with hot water, and the filtrate evaporated to a small bulk; on acidifying the liquid a volu- minous precipitate of methylterephthalic acid was thrown down. This was collected, dried on rz porous plate, and crystallised from glacial acetic acid, when it separated as an apparently amorphous powder melting at 320-330'. The acid was converted into its methylic salt by means of methylic alcohol and hydrogen chloride in the usual way ; after crystallising from methylic alcohol, it melted a t 73-74', and was found to be identical with the methylic salt of this melting point, already described as methylic methylterephthalate (p. 177). On oxidising yamxylic acid in precisely the same way, an acid was obtained which again melted at 320-330" and yielded a methylic salt melting at 73-74" identical with that obtained from xylic acid. These two experiments prove that this substance, melting at 73-74', must be the methylic salt of methylterephthalic acid.180 BENTLEY AND PERKIN : REDUCTION OF XYLIC ACID, ETC. Preparation of MethyZisophthaZiic Acid by the oxidution of Isoxylic Acid. The isoxylic acid required for these experiments was obtained by the oxidation of paraxylyl methyl ketone (CH,),C,H,*CO-CII,, this being prepared by a met'hod similar to that recomicended by Claus ancl Wollner (Bey., 1885, 18, 1856), namely, by treating a mixture of paraxylene and acetyl chloride with aluminium chloride. Paraxylene (20 grams) was mixed with acetyl chloride (25 grams) and carbon bisulphide (60 grams), and finely-powdered aluminium chloride (26 grams) was gradually added t o the mixture, which was well shaken during the operation. The whole was then heated on the water bath for a few minutes, and afterwards poured on t o ice. The oil containing carbon bisulphicle was extracted with ether, the ethereal solution mashed with sodium carbonate, then with water and dried over calcium chloride ; the ether and carbon bisulphide were then distilled OG, and the residue fractionated under the ordinary pressure. About 7 grams of an oil boiling a t 220-230', and consisting of nearly pure paraxylyl methyl ketone, were obtained, ancl 10 grams of paraxylene were recovered unchanged. Oxichtion of paraxpIg2 naethy2 ketone,-The fraction boiling a t 220-230' was mixed with about 30 grams of dilute nitric acid (I vol. of acid, sp. gr. 1.4, and 3 vols. of water) and heated just to boiling for 2 hours. On cooling the mixture, the oily product solidified to a hard mass ; this was collected, washed with water, a i d boiled for a consider- able time with sodium carbonate solution. The liquid was filtered from the insoluble matter, extracted once with ether and then acidified ; the bulky precipitate thus produced was collected, washed with water, dried on a porous plate, and recrystallised from acetic acid. In this way isoxylic acid, C,H,B'Ie,.C0OI3[Me2 : COOH = 1 : 2 : 41, mas obtained quite pure, melting a t 132'. The oxidation of isoxylic acid with permanganate was conducted in exactly the same manner as described in the case of xylic acid (p. 179), and a dibasic acid was obtained which, as it did not give an anhydride, was evideutly methylisophthalic acid. It melted at 320-330°, and yielded a methylic salt which, after crystallisation from alcohol, melted a t 79-80' ; this was identical with the substance of the same melting point described on page 177 as methylic methylisophthalate. OWENS COLLEGE, MANCHESTER.