444 MELDOLA AND KUNTZEIN: SALTS AND ETHERS OFXLVIII.--Salts und Ethers oj' 2 : 3 ;5-li.initro-$-acetyl-amino ph e n o 1.By RAPHAEL MELDOLA, F.R.S., and HAROLD KUNTZEN.THE marked acid character of the above compound, which was firstdescribed by one of the authors in 1906 (Trans., 89, 1935), is shownby the readiness with which it forms metallic and organic salts.At the same time, the extreme mobility of one of the nitro-groupstends to bring about decomposition of the salts, especially inpresence of excess of base. With organic bases of the nature ofprimary amines, the trinitro-compound, as stated in former papers,readily forms salts, but these pass rapidly into catenation productsand finally into iminazoles, so that the intermediate productgenerally consists of a mixture of the salts of the trinitro-compoundwith those of the catenation product, or of the iminazole, or ofboth.With respect to metallic salts, it was pointed out in a recentcommunication (Trans., 1909, 95, 1381) that these could be safelyprepared by the interaction of the trinitro-compound and salts ofmetals with weak acids. This principle has now been successfullyapplied for the preparation of a number of metallic and organicsalts from the acetates, carbonates, etc., of the respective metals oralkaloids. Details of the mode of preparation of the various saltsare given in the experimental part of this paper, the series describedcomprising those which are sufficiently insoluble in water tocrystallise from the hot concentrated solution on cooling.Thisseries is, however, quite typical, and the research has not beenextended to those more soluble salts (lithium, calcium, etc.) whichcould not be directly isolated in the manner described, but whichcould no doubt be obtained if wanted by the evaporation of theirsolutions in a vacuum at the ordinary temperature.Generat Churucters of the Salts.The metallic salts of trinitroadetylaminophenol are all highlycoloured, red substances, the parent compound being pale yellow.*It is for this, among other reasons, that these salts have been con-sidered of sufficient interest to form the subject of a special studyin view of the large amount of work which has of late years beenbestowed upon the subject of colour in relation to chemical con-stitution. It has already been pointed out that the trinitro-compound is capable of a double '' isonitro- " isomerism (Trans.,* The lead salt alone approaches the free trinitro-compound in colour2 : 3 : 5-TRINITRO-4-ACETYLAMINOPHEh’OL. 4451908, 93, 1662), so that a change in constitution in passing fromthe free compound to the salt might reasonably be postulated.Assatisfactory evidence of such change was most likely to be furnishedby a study of the absorption spectra, and as Dr. J. T. Hewitt hasrecently been dealing with this subject (Trans., 1909, 95, 1755),he has, a t our request, been good enough to photograph the absorp-tion spectra of the free compound and its salts, and his observationsare appended to the present paper.As will be seen from theseresults, there is justification for the belief that in forming a saltthe isonitro-constitution is acquired, and he has further obtainedevidence of the transitory existence of a disodium salt in presenceof excess of alkali. I n connexion with these results, it is of interestto note that the methyl ether described in this paper (2: 3 : 5-tri-nitro-4-acetylaminoanisole) is, to the eye, a colourless substance,and therefore may be presumed to have the same constitution asthe free trinitro-compound.All the metallic salts of trinitroacetylaminophenol now madeknown are very soluble in water. They are beautifully crysta.lline,and contain water of crystallisation which in most cases cannot beexpelled at looo, and at higher temperatures decomposition takesplace.They all deflagrate on heating in the dry state, but notexplosively, the only exception being the cobalt salt, whichdeflagrates more sharply than any of the others.I n forming salts with natural alkalaids, the trinitro-compoundshows marked preferential characters. With brucine and guanidinevery stable insoluble salts separate a t once on mixing solutionsof the trinitro-compound and the base, or by the action of thetrinitro-compound on the acetate of the base. Narcotine forms aless stable salt, whilst carbamide, theobromine, quinine, st’rychnine,cinchonine, and morphine do not give readily isolable salts.Caffeine forms a salt which is interesting as being dissociable inalcoholic solution. If equirnolecular weights of the base and thetrinitro-compound are dissolved in a small quantity of boilingalcohol, the solution, on cooling, deposits at first crystals ofcaffeine, and subsequently a mixture of caffeine and the caffeinesalt, the latter crystallising in yellow, nodular aggregates.Furthercrystallisation of the mixed .crystals leads to the same result, evenwhen excess of trinitro-compound is intentionally added, so thatthe pure caffeine salt could not be isolated. In alcoholic solutionthere appears to be an equilibrium mixture, varying in compositionwith temperature and concentration, of free caffeine, free trinitro-compound, and caffeine salt446 MELDOLA AND KUNTZEN: SALTS AND ETHERS OFAttempt to Prepare an Optically Active Compound Containing anAsymmetric Tervalent Nitrogen. Atom.The main object in studying the salts formed by the trinitro-compound with natural alkaloids was to test a, somewhat plausiblehypothesis which had suggested itself with respect to the possibleasymmetry of the nitrogen atom in the trinitro-compound itself.From the formula of this compound,C,H,O C,H(NO,),*OH-NH-C,H,O\/it will be seen that the nitrogen atom is combined with two acidradicles (acetyl and the substituted trinitro-phenol residue) andone positive atom (hydrogen).I n most of the attempts that havehitherto been made to resolve tervalent nitrogen compounds, thenitrogen atom has been combined with positive radicles, and themolecule as a whole has been basic in character." The negativeresults have in these cases been attributed t o racemisation due tohydrolytic dissociation of the salt or to the temporary assumptionof quinquevalent function by the nitrogen atom.The trinitro-compound under investigation is certainly free from the latterobjection, as it is strongly acid in character, and does not formsalts with acids. On the other hand, disregarding for the presentthe possibility of hydrolytic dissociation, if there is any weightattaching to the hypothesis of mutual attfactions and repulsionsbetween the radicles in a molecule, it might be considered thatthree positive radicles attached to a nitrogen atom would bymutual repulsion f avour the configuration sometimes assigned tosuch compounds (No. I), the " bonds '' being in one plane:CGNc b(1.) (11- 1When two strongly acid radicles and one positive atom are presentit seemed, on this view, that every chance for displacement of the" bonds " would be given, and the asymmehic configuration (No.11)assumed. At any rate, the hypothesis seemed sufficiently plausibleto be worth submitting to the test of experiment. The result, as* By way of exception to this general statement, the attempt by Jones andMillington to resolve methylethylanilinesulphonic acid may be referred to (Proc.Canib. Phil. Soc., 1904, 12, 489).IN/\c bf i ' uIn this w e , also, the result was negative2 : 3 : 5-TRINITRO-~-ACETYLAMINOPHENOL. 447in former cases, was, however, negative. A specimen of the brucinesalt, was prepared by precipitating the trinitro-compound in alcoholicsolution with a semi-molecular proportion of the base.The trinitro-compound recovered from the filtrate and from the salt by decom-position by acid was examined for us by Dr. T. Martin Lowry, butin neither specimen was there any trace of optical activity.Dr. Lowry reports that he made his observations with the acetonesolution of the compound (4 grams per 100 C.C. in 2-dcm. tubes)by means of red (lithium) light. We desire to take this oppor-tunity of expressing our thanks to Dr. Lowry for the assistancethus rendered.Although the result is in this case negative, we propose continuingthe investigation, as there still remains th'e possibility that theasymmetry may exist only while the trinitro-compound and thebase are in combination." To test this point, it will be necessaryto prepare some salt more soluble than the brucine salt, and tocompare its optical activity in some non-hydrolysing solvent withthat of the base with which the trinitro-compound is combined.EXPERIMENTALArnmomim Sdt, C8H,08N4=NH,.Prepared by dissolving the trinitro-compound in a hot con-centrated solution of ammonium acetate and allowing to crystallise.Bright red, spherical aggregates of slender needles.The salt isanhydrous, and undergoes decomposition at about 203O when heatedin a capillary tube:0.0434 gave 8.6 C.C. N, (moist) at 12O and 761.3. N=23.56.C8HQOaNS requires N = 23.10 per cent.Sodium Salt, C8H,08N4Na,3H20.Prepared by dissolving the trinitro-compound in a hot con-centrated solution of sodium carbonate.The salt separates slowlyon cooling in long, transparent, ruby-red prisms. Professor W. J.Pope, who has been good enough t o examine these crystals for us,reports that they "probably belong to the anorthic system. Theacute bisectrix of a large axial angle emerges through the smallend faces; the optic axial dispersion is marked, and the angle forblue is larger than that for red light ":0.1336 gave 17.55 C.C. N, (moist) at 12'2O and 747 mm. N=15*31.0.1132 ,, 0.0220 N%SO,. Na=6-31.C&&OsN4Na,3H20 requires N = 15.47 ; Na = 6-36 per cent.* As bearing on this point, see a paper by Pope and Harvey (Trans., 1901, 79,837)448 MELDOLA AND KUNTZEN: SALTS AND ETHERS OFOn heating in the water-oven, the salt becomes opaque and brick-red in colour, and loses weight owing to dehydration and (possibly)partial decomposition, the loss of weight being somewhat in excessof that required by the 3H,O indicated by the above analyses:0.5486, heated in the water-oven, lost 0*0930 = 16.95 per cent.0.0506 (dried as above) gave 7-75 C.C.N, (moist) at 12O and0.1036 (dried as above) gave 0.0238 Na,S04.C,H,O,N,Na requires N=18.19; Na=7*48 per cent.750.1 mm. N=18*01.Na= 7.45.A lossof 3HiO requires 14.9, and of 3$H,O 17.43 per cent.POtaSSkTh Salt, C&,O8N,I(.This salt has already been described (Trans., 1909, 95, 1381).To the former description we are now enabled to add that ananhydrous salt is formed on long exposure to the air a t the ordinarytemperature :N=17.31.0-0678 gave 10.2 C.C. N, (moist) at 17O and 755.5 mm.0.0928 ,, 0.0244 E2S04. K=ll.81.C,H,O,N,E( requires N = 17-28 ; I( = 12.07 per cent.Barium Salt, (C,H,08N4),Ba,3H,0.This salt was prepared by two methods, first by dissolving thetrinitro-compound in a hot solution of barium acetate, andsecondly, by boiling the trinitro-compound with barium carbonateand water and filtering from excess of barium carbonate. In bothcases the solution deposits the salt on cooling as flat needles of adeep red colour and having a slight metallic reflex in the motherliquor when viewed at an angle. The analyses of the specimensprepared by the barium acetate method were somewhat irregular,and indicate that a more definite salt is given by the other method :0.0550 gave 7.15 C.C.N, (moist) at 19O and 758.5 mm. N'=14*9.0.1099 ,, 13.9 C.C. N,,(moist) at 11'5O and 754.1 mm. N=14.92.0-0770 ,, 0.0238 BaSO,. Ba=18*19.Cl,H,,01,N~Ba,3H20 requires N = 14.71 ; Ba = 18.0 per cent.The salt does not lose weight on drying at looo.Magnesium Salt, (C8H508N4),Mg,6H20.Prepared as above from the trinitro-compound and magnesiumacetate. Bright red prisms. The specimen analysed was twicecrystallised from water :0.1292 gave 17.85 C.C. N, (moist) at 18*5O and 753.4: mm.N = 15.782 3 : 5-TRINITRO-4-ACETYLAMINOPHENOL. 4490-1122 gave 0.0161 Mg,P,O,.The salt does not lose water at looo.Mg = 3.21.C,,H,,,0,,N8Mg,6H,0 requires N = 15-95 ; Mg = 3-03 per cent.Prepared as above from the trinitro-compound and zinc acetate.0.1633 gave 21.1 C.C.N, (moist) at 18O and 760 mm.0.3240 ,, 0.0352 ZnO. Zn=8.73.The salt does not lose weight at looo.Bright red prisms, resembling the magnesium sa,lt :N=14*92.C,,H,,0,,N8Zn,6H20 requires N = 15-05 ; Zn = 8.8 per cent.Cadmzhrn SaZ6, (C,H,O,N4),Cd,6H,O.Prepared by boiling the trinitro-compound with water andcadmium carbonate, filtering from excess of carbonate, and allowingto cool, when the salt crystallises out in red scales:N=14*15. 0.1422 gave 16.9 C.C. N, (moist) at 13O and 765.3 mm.0.1602 ,, 0*0262 CdO. Cd=14.32.C,,H,,0,,N,Cd,6H20 requires N = 14.17 ; Cd = 14.22 per cent.The salt darkens in colour and shows signs of fusion and decom-position when heated in the water-oven.Prepared as above from the trinitro-compound and nickelcarbonate.The salt separates as dark brick-red prisms :0.160 gave 22.3 C.C. N, (moist) at 14O and 745.8 mm. N=16*08.0.2460 ,, 0.0260 NiO. Ni=8*31.The salt undergoes no change a t looo.C,,H,o0,GN,Ni,4H,0 requires N = 15-99 ; Ni = 8.37 per cent.It deflagrate; somewhatsharply when heated in the dry state.Prepared as above from the trinitro-compound and cobalt0.1899 gave 24.2 N, (moist) a t 1l0 and 764.3 mm.0.1714 ,, 0.0135 Co. C0=7.88 per cent.C,,H,,0,,N8Co,6H20 requires N = 15.2 ; Co = 8.1 per cent.This salfi differs from the nickel salt, not only in the quantity ofwater of crystallisation with which it combines, but also in thereadiness with which it parts with this water. When heated inthe water-oven, the salt becomes of a dark brick-red colour, andcarbonate.Crystallises in bright red prisms :N=15.24450 MELDOLA AND KUNTZEN: SALTS AND ETHERS OFloses practically all its water. The anhydrous salt becomeshydrated, and changes to a bright red colour on moistening withwater :0.5032, dried in water-oven, lost 0.071.0.0529 (dried as above) gave 7-85 C.C. N, (moist) at 9O and0.0446 (dried as above) gave 0-0042 Co.This salt, on heating, deflagrates more sharply than the nickelR,O = 14.11.C16H10016NsCo,6H,0 requires H,O = 14.66 per cent.749.6 mm. N=17.6.Co=9.44.C16H,,016N8Co requires N = 17-82 ; Co = 9.38 per cent.salt.Manganese Sat t , (C8H,'08N,),Mn,4H20.Prepared m before from the trinitro-compound and manganese0.2152 gave 29.2 C.C.N, (moist) at 12.3O and 749.9 mm. N= 15.9.0.1266 ,, 0.0132 Mn304. Mn=7.51.C16H,,0,,N8Mn,4H,0 requires N = 16.08 ; Mn = 7.89 per cent.The salt darkens and shows signs of fusion and decompositioncarbonate. Dull red, rhombohedral prisms :when heated in the water-oven.Copper Salt, (C8H;08N4)2Cu,4H,0.Dull brick-red, transparent, rhombohedral prisms :Prepared from the trinitro-compound and copper carbonate aa0.1273 gave 17.1 C.C. N, (moistj at 12O and 745.1 mm. N=15*63.0.1709 ,, 0.0190 CUO. Cu=8.89.The salt becomes light brown and opaque on heating in theComplete dehydration couldbefore.C16HloO16NsCu,4H20 requires N = 15-88 ; Cu = 9.0 per cent.water-oven, and partly loses its water.not be effected without decomposition :0'5247 lost 0.0528.H20 = 10.06.C,,H,,O,,N8C'u,4H,O requires H,O = 12 per cent.The salt thus dried gave the following results on analysis:0.0874 gave 13.1 C.C. N, (moist) at 12O and 755 mm.0.0844 ,, 0.0104 CuO. Cu=9.84.N = 17-68.C,6~,,0,,N8Cu requires N = 17-69 ; c u = 10.03 per cent.Lead S d t , (CsH,0sN,),Pb,3H20.Prepared from lead carbonate and the trinitro-compound asabove. Flat, orange needles 2 : 3 : 5-TRINITRO-4-ACETYLAMINOPHENOL. 4510-0948 gave 10.75 C.C. N, (moist) a t 11'5O and 752.2 mm. N = 13.36.0.1428 ,, 0.0525 PbSO,. Pb=25*16.This salt does not lose water at looo.C,,H,,0,,N,Pb,3H20 requires N = 13-48 ; P b = 24.89 per cent.Thtalk6m Salt, C8H;O&?*T1.From the trinitro-compound and thallium carbonate by the same0.0779 gave 7.6 C.C.N, (moist) at 7'7O and 741.4 mm.method. Brick-red, fern-like leaflets :N = 11.52.0.1577 ,, 0.0770 TlCI. T1=41.60.C,H$O,N,Tl requires N;= 11.46 ; T1= 41.72 per cent.The salt undergoes no change a t looo.Silver S d t , C8H,~,N,Ag,3H2O.The salt was prepared from the trinitro-compound and silvercarbonate by the method described above. It generally crystallisesin red prisms, but sometimes separates as orange-red scales. Theanalysis of the salt offered considerable difficulty, as the free trinitro-compound appears to interfere with the precipitation of the silverchloride. Good results were only obtained when the organic matterwas destroyed by heating the salt in a sealed tube with fumingnitric acid for three hours at 190°.The water could not be com-pletely expelled by heating in the water-oven; at this temperaturethe salt darkens, and, on long heating, shows signs of decomposition.About 102O the salt deflagrates :0.0622 (air dried) gave 6.8 C.C. N, (moist) a t 14O and 744.2 mm.0.2996 (air dried) gave 31 C.C. N, (moist) a t 12O and 768 mm.0.2570 gave 0.0820 AgCI. Ag = 24.02.C8E;08N,Ag,3H,0 requires N = 12.53 ; Ag = 24.14 per cent.A specimen dried in the water-oven gave the following results :0.0682 gave 8.2 C.C. N, (moist) at 14O and 730.5 mm.0.0564 ,, 0.0201 AgCl. Ag=26*82.C,H;O,N,Ag requires N = 14-26 ; Ag = 27.46 per cent.The salt is not only readily soluble in water, but also in methyland ethyl alcohols, and by its means the methyl ether has beenprepared, and is described in the present paper.N = 12'59.N = 12.38.N=13*58.This salt was prepared by adding to a hot saturated solution ofthe alkaloid in absolute alcohol an equimolecular proportion of theVOL. XCVII.H 452 MELDOLA AND KUNTZEN: SALTS AND ETHERS OFtrinitro-compound dissolved in the same solvent. It was alsoobtained by dissolving the trinitro-compound in a hot aqueoussolution of brucine acetate and allowing to cool. I n both cases thesame salt is obtained. It is practically insoluble in alcohol, so thatthe trinitro-compound might be used as a precipitant for the base.It can be best purified by dissolving in hot glacial acetic acid,which does not appear to decompose the salt, diluting the solutionwith water, and allowing to crystallise.It separates, on cooling,in rosettes of deep reddish-brown needles, which appear ruby-red bytransmitted light. The salt melts with decomposition at 222-224O :0.2180 gave 0.4366 CO, and 0.0906 33,O. C = 54.62 ; H = 4.89.0.1322 ,, 0.2648 CO, ,, 0.0572 H,O. C=54.62; H=4*80.0.1268C31H32012N6 requires C = 54.68 ; H =4*74 ; N = 12.36 per cent.I n connexion with this salt it is of interest to note that brucine,, 13.7 C.C. N, (moist) at 21° and 75'7.5 mm. N=12.26.also forms a very insoluble picrate.This salt was prepared by dissolving guanidine carbonate indilute acetic acid and saturating the hot solution with the trinitro-compound, the latter being added in small portions in the solidstate. The salt crystallises out on cooling in dark brown needles,appearing ruby-red by transmitted light, and having a bronzy lustrewhen seen by reflected light.Purification was effected bycrystallisation from hot water, in which the salt dissolves withan orange colour:0.1781 gave 44.55 C.C. N, (moist) a t 20'5O and 754.1 mm. N=28.32.0.2170 ,, 0.2502 CO, and 0.0650 H,O. C=31*44; H=3.32.C9H,,0,N, requires C = 31.29 ; H = 3-21 ; N = 28.41 per cent.The salt is quite stable; the melting point is 227O, and no lossof weight takes place a t looo.Nar co tine Sat t , C,H@,N,, C22H,307N.This salt was prepared from the trinitro-compound and the acetateof the base, or by mixing alcoholic solutions of the components inthe usual way. Some difliculty was experienced in obtaining a pureproduct owing to the tendency of the salt to dissociate whencrystallised from alcohol. The most satisfactory results on analysiswere given by specimens crystallised from alcohol in the presence ofa slight excess of the base:0.0988 gave 0.1860 CO, and 0.0378 H,O.C = 51-34 ; H = 4.35.0.1348 ,, 0.2552 CO, ,, 0.0504 H20. C=51*63; H=4.162 : 3 : 5-TRINITRO-4-ACETYLAMINOPHENOL. 4530.0358 gave 3.1 C.C. N, (moist) a t 1 2 O and 761.3 mm.The salt crystallises in flat, yellow needles, melting a t 193-194O.N=10*30.C30H29015N5 requires C =51*48 ; H = 4.19 ; N = 10.02 per cent.Although, for reasons stated in the introductory portion of thispaper, the pure salt could not be isolated, the analytical resultsall pointed to the above formula. The salt, freed as far as possiblefrom admixed caffeine, crystallises in rosettes of yellow needles,melting at about 177O.Dr.J . T . Hewitt's Observations o n the Absorption S p e c t w m ofTriicitroace t ylaminoph en0 I and i t s Salts.The pale yellow trinitro-compound dissolves both in water and96 per cent. alcohol, with an intense yellow colour, inclining toorange. Whilst this orange shade is deepened by the addition ofdilute sodium hydroxide or sodium acetate, the colour in the lattercase soon reaching a maximum, hydrogen chloride has an oppositeeffect, comparatively small quantities causing the solution to assumea pale yellow tint comparable with that of the solid compound. Thisresult is quite in accordance with the fact that trinitroacetylamino-phenol is a fairly strong acid, and necessitates the addition of ahighly ionised acid, such as hydrogen chloride, to the alcoholicsolution if the spectrum of the non-ionised nitrophenol is to beobserved (compare Buttle and Hewitt, Trans., 1909, 95, 1755,e t seq.).*In these circumstances the principal feature of theabsorption spectrum was a band having its head at an oscillationfrequency of about 2800, and not differing very markedly in thisrespect or in its persistency from that observed by Hartley andHuntington in the case of o-nitrophenol (see the curves plotted on alogarithmic scale by Baly, Edwards, and Stewart, Trans., 1906,89, 519).When an alcoholic solution to which a dilute alcoholic solutionof sodium acetate has been added until no further intensificationof the orange shade is noticeable (solution of the monosodium salt),a spectrum is obtained which in one respect shows markedsimilarity to that observed by Baly and his co-workers (Zoc.cit.) inthe case of the sodium salt of o-nitrophenol, and by Buttle andHewitt (Zoc. cit., p. 1756) with that of 2 : 6-dinitrophenol, the bandof slowest vibration having its head a t an oscillation frequency of2250. Hence it seems justifiable to conclude that, as in the caseof o-nitro- and 2 : 6-dinitro-pheno1sy formation of the sodium salt isaccompanied by an alteration in structure, the trinitroacetylamino-aH H 454 MELDOLA AND KUNTZEN: SALTS AND ETHERS OFphenol and its sodium salt being related in the sense of thef ormuk :OH 0jq:NO,Nn {yo2 and NO/[ [NO2NHAc NH- AcNO2\/NO2 \/On the physical side there seems but little doubt that a radicalalteration in absorption on formation of a derivative is accompaniedby a radical alteration in structure, although it is well to keep inmind that the only direct chemical evidence bearing on the assumedchange of structure on salt formation (or ionisation) in the caseof the nitrophenols depends on the isolation of highly colouredaci-esters by Hantzsch and Gorke (Ber., 1906, 39, 1073).If to a solution (aqueous or alcoholic) of the trinitro-compound,alkali be added in excess, a deep purple-red colour is produced;such solutions are, however, unstable, the decomposition makingitself manifest by a.deposition of sodium salt ( 1 nitrite) when asolution in 96 per cent.alcohol is examined, and a, change of colourtowards yellow. Despite this inconvenience, an attempt was madeto photograph the absorption spectrum of a freshly prepared solu-tion, which very possibly contains a disodium salt: the persistencyof the colour band was very similar to that found for a, monosodiumsalt, but the head lies at an oscillation frequency of about 2040.The curves in the figure were obtained with solutions made upin the following manner.TrinitroacetyZaminophenoZ (full curve).-A N / 1000-solution wasprepared from 0.0286 gram and 5 C.C. of fuming hydrochloric acid,made up to 100 C.C. with 96 per cent. alcohol. A N/lO,OOO-solutionwas prepared by diluting 10 C.C. of the N/lOOO-solution and 5 C.C.of fuming hydrochloric acid to 100 C.C. with 96 per cent.alcohol.Monosodium Salt (dotted curve).-O*0286 Gram of the phenol witsdissolved in alcohol, a solution of sodium acetate added in at leasttwice the excess of that necessary to produce any further intensifica-tion of the orange shade, and then made up to 100 C.C. with alcohol.5 111111. of this N/lOOO-solution gave the same absorption as 50 mm.of the solution obtained on tenfold dilution (N/lO,OOO). This showsthat any hydrolysis of the salt is inappreciable, as might have beenexpected.Excess of A Zkali.-The solution photographed was obtained bydissolving 0.0286 gram of trinitroacetylaminophenol in alcohol,adding alcoholic sodium hydroxide in excess of that required formaximum development of the purple shade, and making up to100 C.C.with 96 per cent. alcohol. As 5 mm. of such a solution di2 : 3 : 5-TRINITRO-4-ACETYLAMINOPHENOL. 455not give the same absorption as 50 mm. of a N/lO,OOO-solution, onlythe stronger solution was used. The salt containing more than oneatom of metal is evidently strongly hydrolysed on dilution.Methyl Ether = 2 : 3 : 5-Trinitro-4-acety~am~noa?zisole,0- CH,( y o 2NO,, /NO2NH*CO*CH,Attempts to methylate the trinitro-compound by silver oxide andThe methyl iodide led to negative results (Trans., 1909, 95, 1379).Oscillation freqGencies.18 200022 24 26 28 3000 32 34 36 38 4000 42 44 46T?.in~troc6cetnminophenol i?t alcohol, HCl i n excessY Y ,, , , NaC,H,O, in cxcess¶, ,, ,) NaOH in excess..._.._.._._...__...~.....---___________isolation of the silver salt described in this paper has, however,rendered possible the preparation of the above methyl ether and thecorresponding trinitroanisidine. The silver salt in methyl-alcoholicsolution is rapidly decomposed by methyl iodide, even at theordinary temperature, the separation of silver iodide commencingsoon after mixing the solutions and being complete after about halfan hour's heating on the water-bath.Only about 15 per cent. ofthe trinitro-compound undergoes methylation in this process. Thealcoholic solution, after filtration, is evaporated to a small volume,diluted with water, and extracted with an aqueous solution o456 2 : 3 : 5-Tl~INITItO 4-ACETYLAMINOPHENOL.sodium acetate to remove the unmethylated portion.The residue,after crystallisation from alcohol, consists of white needles, meltingat 194O:0.0566 gave 8.9 C.C. N, (moist) at 13.5O and 762.8 mm. N=18*61.C,H80,N, requires N = 18.67 per cent.That the compound has the above constitution is proved byboiling its alcoholic solution for a short time with a little aniline,removing the excess of aniline by dilute hydrochloric acid,anhydridising the dry product by heating with a little aceticanhydride, and purifying the iminazole thus obtained bycrystallisation from alcohol. The compound was identified as themethyl ether (m. p. 205-206O) of dinitrohydroxy-l-phenylmethyl-benziminazole described in a former paper (Trans., 1908,93, 1672) :0.1074 gave 15.9 C.C. N, (moist) a t 1 2 O and 743 mm.C15Hl,0,N, requires N = 17.07 per cent.As the melting point of the acetyl derivative of the trinitro-anisidine recently obtained by Reverdin (Arch. Sci. phys. nut., 1909,27, 396; 28, 381) is quite different from ours, namely, 242O,Reverdin’s compound is no doubt a derivative of the isomeric2 : 3 : 6-trinitroanisidine. This conclusion is confirmed by a com-parison of the trinitroanisidines, which, by the kindness ofM. Reverdin, who has sent us a specimen of his preparation, wehave been enabled to make. Our acetyl derivative is easilyhydrolysed by heating with excess of concentrated sulphuric acidfor a few minutes to about 105O. The solution, when cold, is pouredinto water, and the trinitroanisidine allowed to separate. Aftercrystallisation from alcohol it consists of dull red, glistening scales,melting at 138-139O :N=21*90.C,H,O,N, requires N = 21-71 per cent.The products ofdiazotisation of the new trinitroanisidine will be of special interest,and we propose extending the research in this direction.N=17.17.0.0632 gave 12 C.C. N, (moist) at 14’7O and 747.8 mm.M. Reverdin’s trinitroanisidine melts at 127O.We have much plea,sure, in conclusion, in expressing our thanksto Mr. Arthur S. Wilson, who has rendered us much assistance inthe course of the work.CITY AND GUILDS TECHNICAL COLLEGE, FINSBURY