254 FORSTER AND JUDD: THE TRIAZO-GROUP. PART X l I ,XXVIL-The Triaxo-y-oup. Pwt X I I . Deyivatives oj'para- T~.iaxobeizxaldehyde.By MARTIN ONSLOW FORSTER and HILDA MARY JUDD.AN investigation of the hydroxyphenylazoimides (Trans., 1907, 9 1,855 and 1350) brought to light a characteristic distinction betweenthe potassium derivative of the para-compound and those of the ortho-and meta-triazophenols depending on the change of colour whichtakes place when the solution in water or absolute alcohol is heated ;as a consequence of this treatment, the colourless substance is trans-formed into a dark blue material, which appears to be isomeric with i t sgenerator. The explanation of this behaviour which suggested itselfat the time depends on the possibility of rearrangement having takenplace in accordance with the following formulae :N N NKN f i ' KO*CGHh*N<j -+ O:CG'EI,:NK<* or O:C,H,:N<and although i t was not possible to recognise any substantialdifference between the acyl derivatives prepared from the bluecompound and those obtained from the colourless potassium salt,they were certainly more highly coloured ; moreover, whilst thep-triazophenol liberated from i t by acids appeara to have the generalproperties of the original material, the solid potassium derivativeregenerated by the action of alkali is not colourless, but bluish-green.The absence of any similiar colour change in the cme of ortho- andmeta-hydroxyphenylazoimides has led us t o examine certain para-derivatives of triazobenzene in which the.occurrence of hydrogen inthe group occupying the para-position afforded an opportunity ofrearrangement on the lines indicated above. The condensationproducts from p-triazobenzaldehyde with hydroxylamine, phenyl-hydrazine, and p-amiuophenol might have been expected, underfavourable circumstances, to yield coloured modifications depending ontransposition of hydrogen FORSTEK. AND JUDD: THE TRIAZO-GROUP. PART XII. 255N NNC,H,* N H N : CH*C,H,*N<* -+ C6H5*N : N*CH: CGH4: N H<gI n no case, however, has any indication of a colour change by alkalibeen observed, and although it certainly happens that a chloroformsolution of p-triazobenzaldehydephenylhydrazone ritpidly becomesbrown and ultimately purple when exposed t o light, the observationsof Chattaway (Trans., 1906,89,462) in connexion with the alterationwhich benzaldehydephenylhydrazone undergoes, would suggest corre-spondence betwean the two phenomena.There does not appear to be any tendency on the part of the triazo-group in p-triazobenzaldehyde t o enter into association with thealdehyde radicle in the manner brought to light by Bamberger andDemuth (Ber., 1901, 34, 1309) during an investigation of certainortho-aminobenzaldoximes. These authors found that on diazotisingo-aminobenzaldoxime there is produced indiazoneoxime, isomeric witho-triazobenzaldehyde, into which it passes on treatment with alkali orwarm water ; furthermore, they showed that o-triazobenzaldehyde maybe converted readily into anthranilic acid through anthranil, fromwhich they regenerated it by the combined action of hydroxylamineand oxygen (Zoc.cit., 3874 and 4015) :As the production of anthranil from o-triazobenzaldehyde mustdepend on an opening of the azoimide ring analogous t o that incurredby diazomethane when this compound transforms aldehydes intoketones (Schlotterbeck, Ber., 1907, 40, 479; 1909, 42, 2559), itseemed reasonable to expect a similar transformation in the case ofo-triazobenzonitrile :The substance in question was therefore prepared by the action ofhydrazoic acid on o-cyanobenzenediazonium chloride ; it was foundthat alkali, however, which transforms o-triazobenzzldehyde intoanthranilic acid with loss of nitrogen, leaves the triazo-group ino-trittzobenzonitrile intact, the substance being merely hy drolysed t 266 FOHSTER AND JUDD: THE TRIAZO-GROUP. PART XII.o-triazobenzoic acid.During this process there does not occur anyelimination of the triazo-group in the form of alkali azide, and thesame power of resistance is displayed by the triazo-group in p-triazo-benzaldoxime. These observations are of some interest in connexionwith the effect produced by other groups on the tenacity withwhich the azoimide complex is held by the benzene ring. Noelting,Grandmougin, and Michel (Ber., 1892, 25, 3328) mere the first toshow that a nitro-group in the ortho- or para-position unloosed thetriazo-group, which remained indifferent to the presence of methyl andthe meta-nitro-complex ; a similar relationship was found applicable tothe naphthalene series (Forster and Fierz, Trans., 1907, 91, 1942),with the addition that a heteronuclear nitro-group in the 5 : 1, 8 : 1,5 : 2, and S : 2 positions left the azoimide radicle unaffected by alkali.Furthermore it was shown (Zoc.cit., 1356) that boiling potassiumhydroxide solution eliminates 20 per cent, of hydrazoic acid fromp-hydroxyphenylazoimide without yielding a trace o€ this productfrom the ortho-derivative.We have taken this opportunity to study the relationship of theoximes obtainable from p-triazobenzaldehyde, and find that theproduction and properties of these substances are governed by thesame generalisations which prevail among their better knownanalogues. The anti-aldoxime, melting at 9So, is the direct productof treating the aldehyde with hydroxylamine, and is converted by theaction of hydrogen chloride in dry ether into the 89%-aldoxime, whichmelts at 142O. Tho latter modification is labile, being transformediuto the more fusible isomerido at looo, and on treatment withbenzoyl chloride in pyridine yields the same benzoyl derivative asthat obtained from the anti-aldoxime ; the p-triazobenzaldoximes thusresemble the unsubstituted compounds, from which only one benzoylderivative is obtainable.On consulting the literature we found thatthe benzoyl derivative of benz-anti-aldoxime was produced by directaction of benzoyl chloride (Minunni and Corselli, Gazzettts, 1892, 22,ii, 167), and consequently in presence of hydrogen chloride.Neverthe-less we have failed to obtain an isomeric benzoyl derivative undermodified conditions, the only product arising by the action of benzoylchloride on tbe dry sodium derivative of both benzaldoximes in etherconsisting of the substance melting at 101" described by Minunni andCorselli; benzoylation in dry pyridine or in presence of aqueousalkali was equally unsuccessful. The beozogl derivative of p-triazo-benzaldoxirne is readily decomposed into p-triazobenzonitrile andbenzoic acid, and on attempting to prepare the p-toluenesulphonylderivative, this nitrile was the sole recognisable product. From thesodium derivative oE benz-anti-aldoxime, however, a very unstablep-toluenesulphonyl compouod has been prepared, melting at 88", anFORSTER AND JUDD: THE TRIAZO-GROUP.PART XII. 257undergoing decomposition into benzonitrile at the laboratory tem-peratme ; the 8yn-aldoxime did not yield a p-tolueoesulphonyl deriv-ative, however, even when exactly the same conditions were observed,benzonitrile alone being formed, and it is noteworthy in this con-ncxion that the nitrile was produced from both oximes by the actionof benzenesulphonyl chloride in pyridine (Werner and Piguet, Ber.,1904, 37, 4308). P-NaphthalenesulphonIl choride appears to beincapable of attacking the isomeric benzaldoximes and their p-triazo-derivatives in pyridine a t zero.EX PE B IM ENTAL.The p-Triuxobeneak~oximes, N,*C,H,*C?:H:NOH,The p-triazobenzsldehyde required for the preliminary experimentswas prepared by the action of sodium carbonate on a solution ofdiazotised p-aminobenzaldeh yde to which hydroxylamine hydrochloridehad been added (Rupe and von Majewski, Ber., 1900, 33, 3401), butwe subsequently obtained a more satisfactory product by the followingprocess.Fifty grams of finely powdered p-aminobenzaldehyde were mixedwith crushed ice, stirred with 150 C.C.of concentrated hydrochloricacid, and diazotised with 25 grams of sodium nitrite in the minimumof water; after half an hour some carbamide was added, and theliquid filtered rapidly from tar. The ice-cold diazonium salt, nowclear and orange-yellow, was treated with 30 grams of sodium azide,added gradually in aqueous solution ; brisk evolution of nitrogenoccurred, and the triazobenzaldehyde separated in the form of a palebrown oil, which, after an hour, was extracted with ether, shakentwice with 10 per cent.hydrochloric acid, and dried with calciumchloride, thereby losing much of the brown impurity. The triazo-benzaldehyde weighed 33 grams,p-Triaxobenx-anti-aldoxime was prepared by adding 6 grams ofhydroxylamine hydrochloride to 5 grams of the aldehyde suspendedin 50 C.C. of 15 per cent. potassium hydroxide, the characteristicodour of p-triazobenzaldehyde, which recalls anise, being no longerperceptible after half an hour. Carbon dioxide was then passedthrough the liquid, and the deep brown precipitate recrystdlisedfrom dilute alcohol, forming lustrous plates which melt at 98" :0 1119 gave 33 C.C. N, at 16" and 769 mm.C7H,0N, requires N = 34.56 per cent.The substance is freely soluble in methyl and ethyl alcohols, ethylacetate, chloroform, ether, glacial acetic acid, acetone, or carbontetrachloride ; benzene dissolves it less readily, and petroleumN = 34.80.VOL XCVII.258 FORSTER AND JUDD: THE TRIAZO-GROUP. PARr XII.sparingly, depositing it in flat, pale yellow needles. It is volatilei n steam, and when crystallised from boiling water is practicallycolourless. Ether extracts small quantities from solutions in alkalihydroxide. When exposed to light the substance rapidly becomesdark brown, but there is no change of colour when a solution in50 per cent. potassium hydroxide is boiled, unaltered materialbeing recovered on acidification ; moreover, one gram of the substancewas heated with 20 per cent.aqueous potassium hydroxide duringfour hours on the water-bath without yielding a trace of potassiumazide. Torrents of gas were liberated by concentrated sulphuricacid, which formed a dark brown liquid indifferent towards Fehling'ssolution. Stannous chloride in !hydrochloric acid also sets free gas,yielding a bright red solution, from which a dark reddish-brownprecipitate quickly separated ; this material, probably an azo-compound,is insoluble in benzene, but dissolves i n alcohol, the deep redsolution becoming yellow with alkali and having its red colourrestored by acids.The methyl ether was obtained by the silver oxide method as acolourless oil having a powerful odour of anise, and also arises by theaction of diazomethane on a solution of the oxime in dry ether(Forster and Dunn, Trans., 1909, 95, 425) ; it is readily volatile insteam, and is not dissolved by alkalis.p-Friaxobenx-syn-atldoxime arose in the form of hydrochloride onpassing dry hydrogen chloride into a solution of the ad-oxime(m. p.9 8 O ) in ether. The salt, after being washed with ether, wassuspended in this medium and decomposed with sodium carbonate, theethereal solution of the syn-oxime being then dried by sodiumsulphate and evaporated at the laboratory temperature, The colour-less, crystalline residue was recrystallised from warm benzene, whichdeposited a felted mass of very pale yellow, silky needles :0,0498 gave 14-2 C.C. N, a t 10.5O and 775 mm.C7H,0E4 requires N = 34.56 per cent.The melting point, 142O, is sharp, although the substance decomposes,The oxime is readily soluble in methyl and ethyl alcohols, from whichwater precipitates lustrous, slender needles ; it is freely soluble incold ethyl acetate, chloroform, ether, acetone, or glacial acetic acid,but is only sparingly so in benzene, and almost insoluble in lightpetroleum. It resembles the more fusible isomeride in being verysensitive to light, volatile in steam, and extractable by ether from itsalkaline solutions.Unlike the sp-modification of benealdoxime andits nitro-derivatives, the triazo-compound appears to resist anytendency to undergo transformation into the anti-form on recrystalli-aation, and must be heated for several minutes at 100° before theN=34*64FORSTER AND JUDD: THE TRIAZO-GROUP.PART XII. 259isomeride is regenerated ; the syn-oxime may even be recoveredunchanged from a solution in 10 per cent. potassium hydroxide whichhas been heated in steam during fifteen minutes.Action, of AcyZ Chlorides on, the 0ximes.-On mixing pyridinesolutions containing 1.4 grams of benzoyl chloride and 1.6 grams of theanti-aldoxime, the temperature rose, and the liquid was thereforesurrounded with melting ice during one hour, and afterwards pouredinto cold water ; the oil precipitated in this manner became semi-solidon treatment with dilute acetic acid, and after being rubbed with afew drops of cold alcohol and drained on earthenware, yielded acolourless solid, only moderately soluble in alcohol, from which itcrystallised in minute, lustrous needles melting at 146' withdecom pod t ion :0.1325 gave 23*5 C.C.N, at 1l0 and 758 mm.On proceeding in exactly the same way with the sp-aldoxime, theproduct is identical.It was noticed that if the oil which is precipitated on pouring thepyridine solution into water is allowed to remain in the liquid without.adding acetic acid, the benzoyl derivative disintegrates in the courseof a few days, yielding p-triaxobenzonitrile, a specimen of which, aftercrjstallisation from alcohol, melted at 71°, and mas therefore identicalwith the substance obtained by Rupe and von Majewski (Zoc. cit.,3406) from p-cyanobenzenediazonium porbromide and ammonia.The action of p-toluenesulphonyl chloride in pyridine was alsostudied, and with this agent both oximes were converted into thenitrile. /?-Naphthalenesulphonyl chloride, however, appears to bedevoid of action, urichanged material being recovered in each casewhen the isomeric oximes were treated with this compound inpyridine.N=21*12.CI4H,,O,N, requires N = 2 1.05 per cent.Action of Potacssium Cganide on p-li.iaxobenxakde~~d.I n preparing p-triazobenzaldehyde by the process of Rupe and vonMajewski, we were led to suspect the occurrence of a subsidiary actionindicated by the comparative indifference of the product towardshydroxylamine and phenylhydraaine, As it seemed possible that aportion of the aldehyde had become converted into bis-p-triazobenzoin,we attempted to compare this compound by the action of potassiumcyanide, but it was soon evident that this agent, instead of effectingbenzoin synthesis, had attacked the trinzo-group in the mannerdescribed by Wolff and Lindenhayn (Bey., 1904, 37, 2374), givingrise to the potassium derivative of the corresponding triazen,0 :CuI * C,H,* N :N*N K a CN.s 260 FORSTER AND JUDD: THE TRIAZO-GROUP.PART XII.Ten grams of the aldehyde in 20 C.C. of absolute alcohol wereheated with 2 grams of potassium cyanide in 10 C.C. of water duringhalf an hour under reflux; the liquid deposited a small quantity ofpotassium cyanide as it cooled, and when filtered from this was allowedto evaporate spontaneously. Bright yellow crystals were obtainedembedded in oil, and on recrystallisation from alcohol formed radialgroups of long, pale yellow prisms :0.0435 gave 0.0179 K,SO,.I(= 18.40.C,H,ON,K requires K = 1840 per cent.The salt has not a definite melting point, decomposing at a hightemperature with extensive intumescence ; it is readily soluble inwater, the yellow colour of the solution being destroyed by hydro-chloric acid and restored by alkali.p- Triazobenxclldehyd~p~en ylh ydrasone, N,*c6H4*CH.: N*NH*C6H,.On mixing 4 grams of p-triazobenzaldehyde with 3 grams ofphenylhydrazine, water separated immediately, the liquid became hot,and solidified on cooling ; the product was recrgstallised from methylalcohol, followed by petroleum, which deposited minute, straw-colouredneedles melting at 190' without evolving gas, although steadyeffervescence began at 120' :0-1103 gave 28.2 C.C.N, at 17' and 767 mm.C13HllN5 requires N = 29.53 per cent.The phenylhydrazone is freely soluble in cold acetone, benzene, ethylacetate, and chloroform, the solution in the last named becoming deepbrown and ultimately purple when exposed to light; this colourchange has not been observed in the case of any other medium, andrecalls the observations of Chattaway in connexion with benzaldehyde-phenglhydrazone (Zoc. cit.). Concentrated sulphuric acid develops atransient, intense blue coloration, followed immediately by liberationof gas ; this was also set free by stannous chloride in hydrochloricacid, which ultimately gives rise to a deep reddish-brown precipitate.There is not any change of colour when the substance is heated withconcentrated potash in either alcohol or water.N = 29-95.Five grams of paminophenol hydrochloride in water were mixedwith 5 grams of sadium acetate and shaken with 5 grams of p-triazo-benzaldehyde, when the emulsion was transformed into a, paste ofcrystals ; after recrystallisation from benzene, the substance waFORSTER AND JUDD: THE TRIAZO-GROUP.PART XJI. 261obtained in lustrous, golden-brown, six-sided pln tes, melting a t 175"with vigorous intumewence :0.1378 gave 27.5 C.C. N, a t 14O and 765-5 mm.C,,H,,ON, requires N = 23.53 per cont.The substance is moderately soluble in boiling methyl alcohol, butonly sparingly soluble in boiling benzene, and insoluble in boilingpetroleum ; acetone and et,hyl acetate dissolve it readily when warm,depositing the substance in clear-cut, hexagonal prisms, whilst boilingchloroform dissolves it sparingly.When suspended in water andheated with potassium hydroxide, the deep yellow solution of thepotassium derivative quickly deposits lustrous crystals, but thesparingly soluble product undergoes no colour change when heatedwith concentrated potash, aqueous cr alcoholic. The substance isdecomposed immediately by concentrated sulphuric acid, liberatinggas, which is also set free by stannous cbloride in hydrochloricacid, the deep red solution rapidly depositing a dark reddish-brownprecipitate.N = 23.64.Condensation o j p-Triaxobenxalclehyde with CamphoryZ-~.semicarbcazide.Fonr grams of camphoryl-$-semicarbazide nitrate dissolved in waterwere warmed with two grams of p-triazobenzaldehyde, when thethe liquid became turbid and co!ourless needles separated.Thehighly insoluble product was exhausted with a small quantity ofboiling absolute alcohol, and then dissolred in a large proportion ofboiIing acetone; dilution with water led to the separation of trans-parent, pale yellow, rectangular prisms, infusible below 250°, althoughsuddenly becoming brown a t about 200' :0.1287 gave 0,2948 CO, and 0-0587 H,O.0.1371 ,, 30.S C.C. N, at 1 5 O and 763 mm. N=26*46.C = 62.45 ; H = 5.06.C,,H,,O,N, requires C = 61 -00 ; H = 6.21 ; N = 23-73 por cent.C,,H,,O3N, '3 C=62*11; H x 5 . 1 7 ; N=26*09 ,, ,,The constitution of the product is somewhat obscure.Theanalytical results are not in agreement with the empirical formulaC,,H,,0,N6, that of p-triazobenzaldehyde camphoryl-$-semicarbazone,and are equally incompatible with the anhydride, which, in view ofthe readiness with which camphoryl-$-semicarbazide loses lH,O,might possibly have arisen. The formula indicated is that of big-p-triazobenzoin camphoryl-$-semicarbazone,CHON" :C(C6H,*N,)*CH(C6H,*N,)*OH]CSHl&(()H)-- NR>"O,less lH,O, an expression which gains probability from the infusi-bility of the substance, because the camphoryl-#-semicarbaaone262 FORSTER AND JUDD: THE TRTAZO-GROUP. PART XJT.already described (Trans., 1905, 87, 727) melt bebow 250°, whilst theanhydride of camphoryl-+-semicarbazide is not fused until 280' isreached.A solution containing 0.1745 gram in 25 C.C.of chloroform gaveaD 7'46' in a 2-dcm. tube, whence [a],, 556*3O, whilst a pyridinesolution with 0.2600 gram in 25 C.C. gnvo aD 0'30' for the samelength, corresponding with [a], 24.0'.CNo-T~iaxobe~xonitrile, /-\IS3. \-/The azoimide was produced by the following steps from o-nitro-aniline :NO, NO, NH2 N3\/ \ \/ \/{)NH -+ / j C N -~ ()CN -+ /)CN. IAdopting the modification of Sandmeyor's method described byPinnow nnd Muller (Rer., 1895, 28, 151), we found that the o-nitro-.benzonitrile was more conveniently separated from the uninvitingproduct of treatment with cuprous cyanide by allowing the blackmass to become dry, extracting it with chloroform, and addingpetroleum to the filtered liquid; in this way, 50 grams of o-nitro-aniline furnished 37 grams of o-nitrobenzonitrile in crystalline form.The difficulties which beset the reduction of o-nitrobenzonitrile too-aminobenzonitrile have been already described by Pinnow andMuller (Zoc.cit.; comprtre also Reissert and Grube, Ber., 1903, 42,3710), and we isolated the hydrochloride of the base in the mannerindicated by them, removing the associated hydrochloride of o-amino.benzamide by treatment with cold water.Fifteen grams of o-aminobenzonitrile hydrochloride were simpendedin 80 C.C. of hydrochloric acid and diazotised with 25 grams of sodiumnitrite; by using a larger proportion of hydrochloric acid than thatemployed by Pinnow and Samann (Rev., 1896, 29, 630), the forma-tion of o-dicyanodiazoaminobenzene, which represented 80 per cent.ofthe theoretical amount in their experiment, ma3 almost entirelyavoided, and the filtered diazonium salt gave o-triazobenzouitrile inthe form of a colourless precipitate on adding 20 grams of sodiumazide. The dried product was rezrystallised from petroleum(b. p. 60-SOo), which deposits pale yellow, lustrous leaflets,quickly becoming brown when exposed to light ; it; sinters at 51°,and is completely fused at 58O :09792 gave 27.0 c.c.' N, at 17" and '748 mm. N= 38.92.C,H4N4 requires N = 38.S8 per centFORSTER AND JUDD: THE TRIAZO-GROUP. PART XII. 263The substance is freely soluble in cold chloroform, ethyl acetate,benzene, or acetone, readily so in alcohol, and more sparingly so in hotpetroleum.Although the solid material is odourless, the vapour insteam has a penetrating odour, without suggestion of anise. There isviolent decomposition with concentrated sulphuric acid.Action of A Zcoholic Potash,-An alcoholic solution of o-triazobenzo-nitrile was heated two hours under reflux with potassium hydroxidedissolved in the minimum quantity oE water, Nitrogen was not setfree, and on evaporating the solution, potassium azide could not bedetected in the residue, which was then dissolved in water and treatedwith dilute sulphuric acid. The precipitated o-triazobenzoic acid wasrecrystallised from boiling water, from which it separated in minute,lustrous needles, melting and decomposing at 146O ; analysis gave25.97 per cent.of nitrogen, 25.76 and 34-57 per cent. being requiredby the formulz C7H,0,N, (triazobenzoic acid) and C,H,ON, (triazo-benznmide) respectively, and the acid is therefore identical with theone obtained by Griess from anthranilic acid (Zeitsch. fiir Chem.,1867, 165) and by Bnmberger and Demuth (Bey., 1901, 34, 1337)from o-t riazo ben zal de h y d e.Aotion of AcyZ Chlorides on the I S O ~ ~ W ~ C Benxaldoximes.So far as we have been able to ascertain, the only benzoyl derivativeof benzaldoxime is that described by Minunni and Corselli (Zoo. cit.),who prepared it from benz-anti-aldosime and benzoyl chloride in dryether. I n view of the existence of two acetyl derivatives correspond-ing with the anti- and syn-aldoximes, it appeared worth while toattempt the production of a benzoyl derivative from benz-syn-ald-oxime under conditions excluding the presence of free hydrogenchloride.As a preliminary experiment, the sodium derivative of benz-anti-aldoxime suspended in dry ether was treated with the calculatedamount of benzoyl chloride, the very vigorous action which ensuedbeing checked by surrounding ice and water; after half an hour,anhydrous sodium carbonate was added to destroy any slight excess ofbenzoyl chloride, when the filtered solution deposited the knownbenzoyl derivative melting at 101".On treating the sodium deriv-ative of benz-syn-aldoxime in precisely the same way, the samesubstance was produced, no depression of melting point occurring onadmixture.An attempt to prepare a p-toluenesulphonyl derivative of benz-anti-aldoxime by the same process was unsuccessful; the residue left bythe solvent remained liquid, and decomposed explosively on slightlyraising the temperature, when torrents of gas were liberated, whils264 PYnlAN : ISOQUINOLINE DERIVATIVES.PART I\*.the tarry residue had the odour of phenylcarbylamine. On shakingthe sodium derivative of benz anti-nldoxime with p-toluenesulphonylchloride in 5 per cent. sodium hydroxide during two and a-half hours,a strong odour of benzonitrilo was developed, and a small proportionof a solid substance remained undiseoIved ; after crystallisation fromlight petroleum, this melted a t 8S0, and was prepared more con-veniently by rubbing the sodium derivative with p-toluenesulphonylchloride and wexk sodium carbonate solution. After remainingtwenty-four hours in the ice-chest, tho pasty mass, with the odourof phenylcarbylamine, had become brittle, and, when washed withsodium carbonate and dried in the desiccator, was dissolved in coldbenzene, from which petroleum precipitated it in lustrous, colourlessprisms melting with profound decomposition a t S8O :0.5359 gave 24.6 C.C. N, a t 24' and 755 mm.Cl,HI,O,NS requires N = 5.09 per cent.The substance is highly unstable, even when purified, especially ifkept in a stoppered bottle, undergoing decomposition after forty-eight hours in daylight t o a tarry mass which has the odour of benzwnitrile ; a specimen on a match-glass in a large desiccator protectedfrom light, however, remained many months without alteration.On proceeding in the same way with benz-syn-aldoxime, the onlyproduct was benzonitrile, and on substituting /3 -naphthalenesulphonylchloride for the toluene derivative, there did not appear to be anyactim with either oxime.N = 5.15.ROYAL COLLEGE OF SCIENCE, LOSDON.SOUTH KENSINGTON, S.V7