THE FORMATlON AND REACTIONS OF IMINO-COMPOUNDS. 2261CCXXXVIII. - 17he Formation and Reactions o jImino-compounds. Part XIV. The Formationof a- Hydrindone and its Derivatives.By ALEC DUNCAN MITCIIELL AND JOCELYN FIELD THORPE.SOME time ago (Trans., 1908,93,165) it was shown that P-hydrindoneand some oE its derivatives could be derived from P-imino-a-cyano-hydrindene (11), a substance which is formed in quantitative yieldwhen an alcoholic solution of o-phenylenediacetonitrile (I), contain-ing a trace of sodium ethoxide, is warmed.CH,*CN >CXH --+ C,H,<CH2>C0(1.1 (11.1 f i - ~ y a ~ i n a ~ n e .C6H'SCIF,.CN -+ C6H4<%t&j aH2262 MITCHELL AND THORPE: THE FORMATION ANDThe present paper deals with a similar reaction by which a-hydrindoneand its derivatives can be produced through the imino-compound.a-Hydrindone was originally prepared by Gabriel and Hausmann(Bey., 1889, 22, 2018), and was subsequently investigated byHausmann (ibid., p.2020). Gabriel and Hausmann prepared theketone by condensing o-cyanobenzyl chloride (111) with the sodiumcompound of ethyl acetoacetate, when a compound, which theyconsidered to be ethyl o-cyano-P-phenylpropionate (I V), was formed,and this substance, on treatment with concentrated hydrochloricacid, passed into a-hydrindone (V).(111.) (IV.) (V. )Hausmann (Zoc. cit.) subsequently found that when ethyl sodio-malonate was used in this reaction in place of ethyl sodioacetoacetate,the same substance (IV) was produced.The formation of a-hydrindone from ethyl o-cyano-P-phenylpropionateon treatment with concentrated hydrochloric acid was explained on theassumption that o-carboxy-P-phenylpropionic acid (VI) is first formed,which then decomposes into water, carbon dioxide, and a-hydrindone,thus :(VI.1This is also the explanation advanced by Aschan (Chemie derAZicykZischen Verbindumgen, p. 1028), who remarks that Konig hasshown (Anrtctlen, 1893, 275, 341) that o-csrboxy-P-phenylpropionicacid passes on distillation into a-hydrindone. It must be remembered,however, that the production of a-hydrindone from o-carboxy-P-phenyl-propionic acid in this manner takes place at a high temperature, thatis to say, under :conditions very different from those which convertethyl o-cyano-/3-phenylpropionate into this ketone.It therefore seemedto us unlikely that the mechanism of this reaction, as recorded above,could be correct, and we consequently decided to seek for some otherexplanation more in accordance with the experimental facts.It has been already mentioned that, according to the observation ofHausmann, the condensation of o-cyanobenzyl chloride with both ethylsodiomalonate and ethyl sodioacetoacetate yields the same product,namely, ethyl o-cyano-P-phenylpropionate (IV). It is evident, there-fore, that in the first condensation a carbethoxyl group, and in thesecond condensation an acetyl group, must have been eliminated duringthe process of the condenaation.Neither Gabriel nor Hausmann seems:to have remarked on this, butREACTIONS OF IMlNO-COMPOUNDS. PART SIV.2263in the light of recent investigation on the formation of five-ring imino-compounds, the fact possessed for us some significance.Thus we have always found that when the five-carbon ring is formedthrough the imino-group, the formula of the product does not allow oftwo carbethoxyl groups, a nitrile group and a carbethoxyl group, or anacetyl group and a carbethoxyl group, being attached to the same carbonatom; an example of this, which bears closely on the present instance,being the transformation of the open-chain compound (VII), which isformed by the condensation of ethyl sodiomalonate and ethyl 1-cyano-cydopropane-1-carboxylate into ethyl 2-iminocyclopentane-1 : 3-dicarb-oxylate (VIII) (this vol., p.1002).(VII.)QH,*CH(CO,Et) CH2* UH( C0,>C:NH.The elimination of a carbethoxy-group in Hausmann’s condensationsuggested therefore the closing of the five-carbon ring, in whichcase the reaction would have proceeded as follows :(VIII.)or in the case of ethyl sodioacetoacetate as follows :C,H4<E&C1 + CHAcNa*CO,Et -+CNcAH4<CXI,*CHbc*C0,Eti i- NaC1*+ EtOH -+ CNC6H4%H,*CHAc*C0,EtC 6 H 4 < ~ & ~ ~ > C H * C O z E t + Me*CO,Et.In these circumstances the compound described by Gabriel andHausmann as ethyl o-cyano-P-phenylpropionate would be ethyl 1-imino-hydrindene-2-carboxylate (IX), and its transformation into a-hydr-indone by the action of concentrated hydrochloric acid could be readilyexplained thus :C,H4<~&~H&CH*C0,Et -+ C6H4<Co->CH*C02Et CH2 --2264 MITCHELL AND THORPE: THE FORMATION ANDInvestigation proved that this view of the formation of a-hydrindonewas correct, and that the compound described as ethyl o-cyano-p-phenyl-propionate behaved in every way as an imino-compound of formula (IX).Before this fact could be definitely proved it was necessary, however,t o thoroughly study the condensation of o-cyanobenxyl chloride withthe sodium compounds of both ethyl malonate and ethyl acetoacetate,and as, at the same time, it was thought advisable, for reasons giveniater, to investigate the corresponding condensation with ethyl sodio-cyanoacetate, it is best, for the sake of comparison, to describe thethree condensations separately.(1) The Condensation of o-Cyanobenxyl Cht?oride with the XodiumCompound o j EthylMa1onate.-In effecting this condensation, Hausmannused equivalent quantities of sodium dissolved in alcohol, and of thetwo reacting substances.Hs found that, mixed with the chief productof the condensation (the so-called ethyl o-cyano-P-phenylpropionate), aconsiderable quantity of a product melting at 8 6 O was also formed.This he showed to be ethyl di-o-cyanobenzylmalonate (X),(CN*C,H,*CH,),C(CO,Et),,(X. )which had been formed by the condensation of two molecules of0-cyanobenzyl chloride with one molecule of ethyl malonate.Hausmann separated the two products by treating the mixture withcold concentrated hydrochloric acid, in which the supposed ethylo-cyano-p-pheny lpropionate dissolved, and could be obtained on mixingthe hydrochloric acid filtrate with water.Now it is obvious that the formation of the derivative (X) musthave taken place in the following manner :(1) C?N*C),H,*CH,Cl + CHNa(CO,Et), -3CN~C,H,~CH,~CH(CO,Et), + NaCI.(2) CN*C6H,*CH2*CH(C0,Et), + CHNa( CO,Et), -+ON*C,H4*CH2GNa(C0,Et), + CH,(CO,Et),.(3) CN*C6H,*CH,*CNa(C0,Et), + CN*C6€€,*CE,CI -+( CN*C6H40CH,),C(C0,Et)2 + NaCl.That is to say, the initial condensation product of o-cyanobenzylchloride and ethyl sodiomalonate must be the normal product ofthe formula CN*C,H,*CH,=CH(CO,Et),, and the elimination of acarbethoxyl group must therefore have taken place subsequent toits formation,Our previous experiments on this point show that the elimination ofthis group is always effected by the action of free sodium ethoxide, orof some sodium derivative dissociating in alcoholic solution, and henceit seemed to us likely that by preventing, so far as possible, the presencREACTIONS OF IMINO-COMPOUNDS.PART XTV. 2265of excess of the sodium derivative of ethyl malonate during the processof the condensation the elimination of this group might be avoided.This proved to be the case, for it was found that when an alcoholicsolution of ethyl sodiomalonate containing a slight excess of ethylmalonate was added slowly to a hot alcoholic solution of o-cyanobenzylchloride, the product of the reaction did not become solid on beingpoured into water, but remained as a heavy oil at the bottom of theliquid.This oil proved to be the normal condensation product, namely,ethyl o-cyanobenzylmalonate, CN*C,H,*CH,*CH(CO,Et),, and it wasfound that when an alcoholic solution of this substance containing atrace of sodium ethoxide was warmed, a carbethoxyl group was at onceeliminated as ethyl carbonate, and the product described by Hausmannas ethyl o-cyano-P-phenylpropionate was formed.Subsequent investi-gation proved conclusively that this product is ethyl 1 -iminohydrin-dene-2-carboxylate (IX), and that its formation in the manner des-cribed above is represented by the equation :UX.1The proof of the constitution of this substance is as follows : Whena solution of the imino-compound in alcohol is mixed with rather morethan the quantity of concentrated hydrochloric acid necessary tohydrolyse the C:NH-group to carbonyl and the solution is warmed,ammonium chloride separates, and the solution on dilution yields an oilwhich boils a t 185'/20 mm., and which gives in alcoholic solution anintense violet coloration with ferric chloride.This oil, which is withoutdoubt ethyl 1 -hydrindone-2-carboxylate (XI),gives a well defined phenylhydrazone and semicarbazone, the samephenylhydrazone being also formed from the imino-compound (IX)when it is boiled in acetic acid solution with phenylhydrazineacetate.The imino-compound is, as Gabriel and Hausmann showed, readilysoluble in concentrated hydrochloric acid, and is precipitated on addingwater. It is not, however, completely unchanged by this process, sincea quantity, depending for amount on the length of time it is left incontact with the strong acid, is converted into the ketone. The processof conversion at the ordinary temperature is very slow, and the usualmethod adopted in other cases, of pouring the concentrated hydro-vm. XCVII.7 2266 MITCHELL AND THORPE: THE FORMATlON ANDchloric solutiori, into hot water and cooling, converts only a smallquantity of the imino-compound into the ketone. It is evidenttherefore that ethyl l-iminohydrindene-2-carboxylate is a tautomericamino-imino-compound reacting in the two forms :but thiit it has only a short imino-phase.E thy1 1-hydrindone-2-carboxylate (XI) is readily soluble in diluteaqueous potassium hydroxide, but i t is only slowly extracted from itssolution in ether by means of aqueous sodium carbonate solution.Both the potassium and sodium salts are sparingly soluble inexcess of the alkali, and can be readily isolated in a pure condition.When the potassium salt (XII), either in the soluble or insoluble form,is boiled in alcohol with methyl iodide, the C-methyl derivative (XIII)is obtained as sole product, and no trace OF the corresponding O-methylether could be detected :c6H4<Et7cK* CO,Et -+ C,H4<Egi>CMe*C0,Et(XII.) (XIII.)The C-methyl derivative readily yields the corresponding 2-methyl-l-hydrindone when distilled in a current of steam from dilute sulphuricacid.a process which is complete, owing to there being no tendency forthe P-alkyl derivatives of a-hydrindone to undergo intramolecular con-densation analogous to the formation of anhydro-bis-a-hydrindone(Trans., 1897, 71, 241) from a-hydrindone. The direct formation ofethyl l-iminohydrindene-2-carboxylate in the original condensation is,we find, best effected by working in the manner described above untilthe reaction is complete, and then, by adding excess of sodium ethoxide,to convert the open-chain compound into the hydrindene ring.Thecompound CN*C,H,*CH,*C(CO,Et),*C A,*C,H,=CN, which is formed bythe condensation of two molecules of o-cyanobenzyl chloride with onemolecule of ethyl malonate, is so readily produced that if the conden-sation is effected in the usual way, the product consists for the most partof the more complex derivative.(2) The Condensation of Ethyl Xodioacetoacetate and o-CyanobenxylChZoride.-It is hardly necessary in this condensation to take anyprecautions t o prevent the formation of the derivativeCN*C,H,*CH,* CAc( C0,Et) CH2*C,H4 *CN,as under the ordinary conditions very little of it is formed.If, how-ever, the method described in the experimental portion is used, thenormal product (XII) can be readily isolated, and this compound oREACTIONS OF IMINO-COMPOUNDS. PART XTV. 2267treatment with a trace of sodium ethoxide passes quantitatively intoethyl acetate and ethyl 1-iminohydrindene-2-carboxylate, thus :+ EtOH -+ CNCaH4<CH,*CHAc*C0,EtC6H4<E&y>cH. C0,E t + Me C0,Et.For the rapid preparation of this imino-compound in quantity thiscondensation produces the best results, the best method being to addan alcoholic solution of ethyl sodioacetoacetate containing a slightexcess of sodium ethoxide to a hot alcoholic solution of the chloride.( 3 ) The Condensation of Ethyl Sodiocyanoacetate and o-CyanobenzylChloride.-The main object of investigating this condensation was tocompare the derivatives of a-hydrindone with those of /?-hydrindone inorder to ascertain whether the phenomenon of '' steric inhibition "which was so marked in the case of the P-compound substituted inthe a-position applied also to the a-compound substituted in theP-position.It has been found as regards the derivatives of P-hydrindone and itsimino-derivative (Trans., 1908, 93, 165), that the presence of certaingroups on the a-carbon atom causes the compounds to react as trueamino-derivatives, whereas certain other groups on this carbon atomcause the imino-form to be stable, and that this occurs irrespective ofthe acidity of the groups which we have hitherto shown to be thedetermining factor in deciding the amino- or imino-structure of acompound of this type.Thus the compound of formula (XIIIa)evidently had the imino-structure, or rather exhibited amino-imino-tautomerism with a long imino-phase, whereas when the cyano-groupwas displaced by the less negative carbethoxyl group, the compound,instead of showing amino-imino-tautomerism with a longer imino-phase, as it should have done, behaved as a true amino-compoundof formula (XIV) :(XIIIa.) (XIV.)It seemed of interest thereFore to compare the corresponding deriv-atives of a-iminohydrindene of formuh (XV) and (XVI)(XV.1 (XVI.)in order that a direct comparison might be made between the twoseries.The condensation of o-cyanobenzyl chloride with the sodium7 1 2268 MITCRELL AND THORPE: THE FORMATION ANDcompound of ethyl cyanoacetate, if carried out in the usual may, leadsto the formation of the compoundCN *C,H,* CH,*C(CN) (CO,Et)* CH,* C,H,* CNto the extent of 80 per cent.of the theoretical amount, and only asmall quantity of ethyl 1-imino-2qanohydrindene (XV) is produced.If, however, the precaution is taken of adding an alcoholic solutionof the sodium salt of ethyl cyanoacetate t o a hot alcoholic solutionof o-cyanobenzyl chloride, a considerable yield of the normal condensa-tion product (XVII) can be obtained, and this substance on treatmentwith sodium ethoxide passes at once into 1-imino-2-cyanohydrindeneand ethyl carbonate in accordance with the scheme :+ EtOH -+ ONC6H4<CH,*CH( CN) C0,Et(XVII.)C,H,<Z&:E?>CH*CN + CO(OEt),.A comparison of 2-imino-1 -cyanohydrindene (XIII) with l-imino-2-cyanohydrindene (XV) showed that these compounds did not possessany essential points of difference beyond what was to be expected fromtheir structure.Thus the imino-group of 1-imino-2-cyanohydrindeneis between the negative phenylene group and the nitrile group,whereas in the case of 2-imino-1-cyanohydrindene the negativeinfluence of the phenylene group would be less pronounced.Conse-quently it might be anticipated that whilst both compounds wouldbe tautomeric amino-imino-compounds, the 1-imino-derivative wouldpossess a longer amino-phase than the 2-imino-derivative.This anticipation is borne out by the experimental facts, forwhereas 1 -imino-2-cyanohydrindene, with its long amino-phase, dis-solves readily in concentrated hydrochloric acid, forming a salt which,on the addition of water, is dissociated regenerating the amino-compoundmixed with only a small quantity of the corresponding ketone, 2-imino-1 qanohydrindene does not dissolve in concentrated hydrochloric acid,but when warmed with the acid is converted almost completely intothe ketone. I n other words, the rate of hydrolysis of the 2-imino-derivative is very much quicker than that of the 1-imino-derivative.The two compounds therefore serve as admirable ex.amples of theinfluence of negative groups on the predominance of the amino- orimino-phase i n compounds exhibiting amino-imino-tautomerism.When the two carboxylic esters of formuh (XVIII) and (XIX)(j0H4<~&~f~>C*C02Et and C0H4<CH(Co2Et)>C: CH2- NH(XVIII.) (XIX.)are compared, a marked difference is at once apparent.T t has been already shown that ethyl 2-iminohydrindene-l-ca1-bHEACTIONS OF IMLNO-COMPOUNDS.PART X LV. 2269oxylate (XIX) is a true amino-compound, which can be hydrolysedto the corresponding acid by means of hydrochloric acid without thenitrogen group being affected. Subsequent experiments showed thatthis behaviour was exhibited by other compounds having groups oflarge molecular volume attached to the 1-carbon atom, and hence itwas suggested that the presence of a group of more than a certainvolume did not permit of the attachment of the hydrogen atom to the1-carbon atom.In other words, the compound ceased, under theseconditions, to react in one of its tautomeric forms.The examination of the behaviour of ethyl l-iminohydrindene-2-carboxylate (XVIII) under similar conditions showed that, althoughthe substitution of the carbethoxyl group for the nitrile groupincreased the amino-phase of the compound, that is to say, increasedthe length of time required for the hydrolysis of the compound to theketone, yet it still exhibited well-defined amino-imino-tautomerism, andits hydrolysis to the ketone by acids was always completely effected,It is, of course, evident that as the carbethoxy-group possesses lessnegative properties than the nitrile group, the reverse should be thecase, and it therefore follows that there must be a certain degreeof steric hindrance attaching to the 1-carbon atom, but to a very muchless extent than to the %carbon atom.An explanation of this factis aEorded by the consideration that the 1-carbon atom of ethyl2-iminohydrindene-1-carboxylate is attached t o both the phenylenegroup and the carbethoxyl group, whereas the 2-carbon atom of ethyl1-iminohydrindene-2-carboxylate has only the carbethoxyl attachedto it.When the oxygen derivatives (XX) and (XXI) are compared, certainpoints of difference are also observed. Thus it has: been shown that1-cyano-2-hydrindone (XXI) yields a methoxy-derivative (XXII) whenalkylated by the usual etherifying agents, but that when methylatedby means of sodium methoxide and methyl iodide it yields the C-methylderivative (XXIII).With 2-cyano-1 hydrindone (XX) the increasednegative character of the compound due to the proximity of thephenylene group is apparent, and the action of sodium methoxideand methyl iodide leads t o the formation of the methoxy-derivative(XXIV) only22'10 MITCHELL AND THORPE: THE FORMATION ANDAs considerable quantities of the compounds containing twoequivalents of o-cyanobenzyl chloride had accumulated during thepreparation of the simpler products formed in these condensations, weinvestigated their properties.The cyano-derivative (XXV) formedin the ethyl cyanoncetate condensation is readily hydrolysed by dilutealkali, yielding an alkali salt, from which the acid (XXVI) is obtainedby the action of mineral acids, When heated at lSOo, this substanceCN*C,H,*CH,*C(CN)(Co,Et)* CH,*c6H4*cN(XXV.)ONm C,H,* CH,* c( CN) (C0,H) cH2*c6H, * ON(XXVI.)CN*C6H,*cH2*CH( CN)*CH,*C,H,*CN(XXVII.)CO,H*G,H,*CH,* CH( CO2H)*CH2*C6H4*CO2H(XXVIII.)eliminates carbon dioxide and passes into the trinitrile (XXVII).When completely hydrolysed, the nitrile is converted into the tri-carboxylic acid (XXVIII), a compound which is identical with thatformed by the complete hydrolysis of the condensation products(XXIX) and (XXX) formed in the ethyl malonate and ethyl aceto-acetate condensations respectively.The constitution of these morecomplex compounds is therefore clearly established.CN*C6H, CH2* C( CO,Et),* CH2* C6H4* CN(XXIX.)CN * C6H,* CH,*CAc( C0,Et) *CH,= C,H,- CN(XXX.)EXPERIMENTAL.Etl.?) o-Cyanobenxyt?ma~onate, CN*C6H4*CH,*CH(C0,Et),.This substance can be prepared in quantity by the condensation ofethyl sodiomalonate with o-cyanobenzyl chloride if care is taken not tohave free sodium ethoxide present at any time during the course ofthe condensation. In order to effect this, i t is necessary to add analcoholic solution of the sodium compound of ethyl malonate to a hotalcoholic solution of the chloride, that is to say, in the reversemanner to that usually adopted in these condensations.The con-ditions found most favourable were as folbws : 2.2 grams of sodiumwere dissolved in 50 C.C. of alcohol and mixed with 16 grams of ethylmalonate, the warm solution being then slowly added to a hot solutionof 15 grams of o-cyanobenzyl chloride dissolved in the requisiteamount of alcohol. The reaction was allowed t o proceed by its ownheat, and was completed by heating on the water-bath for five minutes,when the solution was found to be neutral. Water was then added,and the alcohol and unchanged ethyl malonate separated by distillatioREACTIONS OF IMINO-COMPOUNDS. PART XIV. 227 1in a current of steam, the heavy, non-volatile oil being subsequentlyextracted by ether.The residue which remained after evaporating theether yielded a large fraction, boiling at 213'/20 mm., consisting of aviscid, colourless oil :0.1958 gave 0.4722 CO, and 0.1080 H,O.C,,H,704N requires C = 65.4 ; H = 6.2 per cent.The ethyl salt is unchanged by cold concentrated hydrochloric acid,and its open-chain structure is proved by the fact that when boiled fora long time with dilute sulphuric acid, i t is slowly hydrolysed, and thesolution on cooling deposits a ci-ystalline acid, which, when recrystal-lised from water, yields long needles, melting at 166'. (Found,C = 61-72 ; H = 5.3.C = 65.77 ; H = 6.13.Calc., C = 61.8 ; H = 5.2 per cent,)The acid is therefore o-carboxy-j3-phenylpropionic acid,C0,H C,H,*CH2* CH,*CO,H.During the fractionation of ethyl o-cyanobemylmalonate a smallquantity 6f lower boiling material was obtained, which solidified aftersome time, and, on examination, proved to be ethyl l-iminohydrindene-2-carboxylate (see p.2273). There was also a small amount of higherfraction boiling at about 300'/20 mm., which also solidified, and provedon investigation to be ethyl di-o-cyanobenzylmalonate (see p. 2280).Ethyl l-lminohydrindene-2-carbox~late, C,H,<:$>CH* C0,Et.This substance may be prepared in one of two ways, namely :(1) By the Action of Alcoholic Sodium Ethoxide on Ethyl o-Cyanobenxyl-ma1onate.-This method, which gives a quantitative yield of the imino-compound, can be carried out as follows : 10 grams of the ethyl saltare diluted with twice its volume of alcohol, and 1 C.C.of a solution of1 gram of sodium in 12 C.C. of alcohol is added. The solution, whichbecomes appreciably warm, is kept for fifteen minutes, when it iswarmed on the water-bath for five minutes and then poured into anequal volume of water. The crystals which separate are thencollected and recrystallised from dilute alcohol.The mother liquor from the crystals was extracted with ether, andthe residue, af t u r evaporating the ether, carefully fractionated underthe ordinary pressure. A fraction boiling a t 126-12'7' was ulti-mately obtained, which analysis showed to be ethyl carbonate.(Found, C = 50.65 ; B = 8.6.(2) By the Direct Condensation of Ethyl Sodiomalonate and o-Cyano-belzzyl Chloride.-This method was that used by Gabriel and Hausmann,but under the conditions employed by them, we find that the greaterportion of the product consists of ethyl di-o-cyanobenzylmalonate, andonly a small yield of the imino-compound can be obtained.It is men-Calc., C = 50.8 ; H = 8.5 per cent.2272 MITCHELL AND THORPE: THE FORMATION ANDtioned later (p. 2279) that the best method for preparing this imino.compound in quantity is by employing the sodium compound of ethylacetoacetate instead of ethyl sodiomalonate in the condensation, becausein that case very little of the di-derivative is formed. A good yield ofthe imino-compound can, however, be obtained from ethyl malonate ifthe conditions described in the preparation of ethyl o-cyanobenzyl-malonate (p.2270) are closely followed, and, as soon as the con-densation mixture has become neutral, a small quantity of alcoholicsodium ethoxide is added, and the heating continued for five minuteslonger. When the product obtained in this way is poured into water, itwill be found that the whole of the dicarboxylic ester has been con-verted into the imino-compound and ethyl carbonate, and that it willdeposit a large quantity of oil which will solidify on scratching. Thissolid, which consists of the imino-compound mixed with some ethyldi-o-cyanobenzylmalonate, can be separated by the method used byGabriel and Hausmann, that is, by dissolving the imino-compound inconcentrated hydrochloric acid and filtering the solution from theundissolved di-derivative.On diluting the filtrate, a certain amountof the imino-compound separates in the crystalline form, but theseparation is by no means complete, as the substance is appreciablysoluble in dilute hydrochloric acid. I n order to obtain the wholeamount, it is necessary to extract the diluted acid solution with ether.Ethyl 1 -iminohydrindene- 2- carboxylute crystallises from dilutealcohol in colourless needles, which melt a t 98'. It gives no colorationwith ferric chloride :0.1923 gave 0.4970 CO, and 0.1060 H,O.C12H1302N requires C = 70.9 ; H = 6.4 per cent.The mother liquor from the recrystallisation of the imino-compoundgives an intense blue coloration with ferric chloride, showing that thetreatment with hydrochloric acid had converted some of it into thecorresponding ketone.When treated in hot acetic acid solution withphenylhydrazine acetate it yields the same hydrazone, melting atlOl*S0, as that derived from the ketone (p. 2273). It is insoluble inaqueous potassium hydroxide, and is only slowly hydrolysed on boilingwith this reagent.Action of Nitrous Acid.-When the imino-compound is dissolved inconcentrated hydrochloric acid and mixed when very cold with excessof sodium nitrite solution, an oil is precipitated which solidifies aftersome time. When this substance is recrystallised from alcohol, it isobtained in brilliant yellow leaflets, which melt at 163" with vigorousdecomposition and charring.C = 70.49 ; H = 6-10,Analysis points to the formulREACTIONS OF IMINO-COMPOUNDS. PABT n v .2273but the position of the nitroso-group in the ring is uncertain :0.1765 gave 0.3559 CO, and 0.0632 H20. C= 54.99 ; H= 3.95.C,,H,,O,N, requires C: = 55.0 ; H = 3.8 per cent.Ethyl I-Nycls.indone-2-cni.60xylccte, C,H,<Co->CH* C02Et.( 3 3 2Ethyl 1-iminohydrindone-2-carboxylate is only very slightly changedwhen its solution in concentrated hydrochloric acid is poured intoboiling water, conditions which completely hydrolpse the correspondingnitrile to the ketone (see p. 2277). The hydrolysis can, however, becompletely effected in the following manner : Ten grams of the imino-compound are dissolved in 50 C.C. of alcohol, and rather more than thecalculated quantity of concentrated hydrochloric acid is added. Thesolution is then boiled for three minutes, during which time a largequantity of ammonium chloride separates.It is then cooled andmixed with a large volume of water, when a heavy oil is deposited,which is extracted by ether. The ethereal solution, when freed fromimpurities by washing first with water and then with sodium carbonatesolution, leaves, on evaporation, an oil which distils at 185'/20 mm. asa viscid, colourless liquid. The distillation can only be accomplishedwith emall quantities, otherwise rapid decomposition ensues :0.2169 gave 0.5593 CO, and 0.1168 H,O.EthyE l-hydrindone-2-carboxyZai% gives in alcoholic solution an intenseblue coloration with ferric chloride. It is soluble in dilute aqueouspotassium hydroxide and in cold alkaline carbonate solutions, but itcannot be extracted from its solution in ether by shaking with thesereagents.When excess of aqueous potassium hydroxide is added to asolution of the ketone in the dilute alkali, a sparingly soluble potassiumsalt separates. When freshly precipitated, this salt is readily solublein both ethyl and methyl alcohol, but if the solution in either of thesesolvents is kept, colourless, silky needles separate, which are verysparingly soluble in hot alcohol :0,3192 gave 0.1127 K2S0,. K = 15-83.C,,H,,O,K requires K = 16.1 per cent.It is probable that the salt whenfirst precipitated contains water ofcrystallisation, although no satisfactory analysis could be made of thehydrated product.The phenylhydraxone, C18H1802N2, is formed when either the ketoneor the imino-compound, dissolved in glacial acetic acid, is mixed with asolution of phenylhydrazine acetate and boiled.It separates as an oilon dilution, and solidifies on scratching. When recrystallised fromalcohol, it forms pale yellow needles, which melt at 1 0 1 * 5 O :C = 70.33; H=5*98.C,,H,,O, requires C = 70.6 ; H = 5.9 per cent2274 MITCHELL AND THORPE: THE FORMATION AND0.1861 gave 0.5001 CO, and 0.1036 H,O.Cl8HI8O2N2 requires C = 73.5 ; H = 6.1 per cent.The sernicarbazone, C13H1503N3, is precipitated when a solution ofthe ketone in dilute alcohol is mixed with an aqueous solution ofsemicarbazide acetate. It crystallises from dilute acetic acid in slenderneedles, which melt and char at 200° :C= 73-30 ; H= 6.13.0,1445 gave 0.3161 CO, and 0.0756 H,O.l-€€~drindone-2-carbuniZide, c 6 H 4 < ~ ~ ~ c H * C O * N H l ? h , is formedwhen the ethyl salt is boiled with an equal volume of aniline for fiveminutes and the solution is poured into dilute hydrochloric acid.Itcrystallises from alcohol in colourless, lustrous plates, which meltat 177" :C=59*66 ; H=5*81.C13Ml,03N8 requires C = 69.8 ; H = 5.8 per cent.0.1560 gave 0.4369 CO, and 0,0754 H20. C = 76.38 ; H = 5.37.C16H,,02N requires C = 76.5 ; H = 5.2 per cent.a-Hydrindone.This ketone is most conveniently prepared by passing a current ofsteam through ethyl 1-hydrindone-2-carboxylate or the imino-compoundsuspended in boiling 10 per cent. sulphuric acid. The ketone passesover with the steam, and solidifies in the receiver. It melted at 41°,and was characterised by conversion into the semicarbazone melting at237O.The formation of anhydro-bishydrindene was not observed, theyield of the ketone being practically quantitative.Ethyl 3-Meth yl- 1 -hydrindone-2 -carboxyZate, ~C6H4<~~~>CH*CO2Et.This compound may be prepared in the following wa.y: Eightgrams of ethyl 1-hydrindone-2-carboxylate are added to a solutioncontaining 1 gram of sodium dissolved in 20 C.C. of alcohol, when acopious precipitation of the sodium salt takes place. Excess of methyliodide is then added, and the mixture is heated on the water-bathuntil the sodium compound, which is practically insoluble in hotalcohol, has all passed into solution. The product is then freed fromalcohol by evaporation on the water-bath, diluted with water, and theoil which is then precipitated is extracted by ether.The etherealextract, after being washed with sodium carbonate solution, is driedand evaporated, when it leaves an oil which distils at 181°/20mm. as amoderately viscid, colourless liquid :0.1848 gave 0.4817 GO, and 0*1848 H20. C= 71.09 ; H = 6.70.C,,H,,O, requires C = 71.6 ; H = 6.4 per centREACTIONS OF IMINO-COMPOUNDS. PART XIV. 2275Ethyl 3-mthyZ- 1 - h ~ d ~ ~ i r t d o n e - 2 - c a r ~ o x y ~ ~ is insoluble in aqueouspotassium hydroxide, and gives no coloration in alcoholic solutionwith ferric chloride.The semicarbazone, C14H1703N3, is precipitated when a solution ofthe ketone in dilute alcohol is mixed with an aqueous solution ofsemicnrbazide acetate.It separates from dilute methyl alcohol inclusters of small needles, which melt at 150' :0.1692 gave 0.3794 CO, and 0.0956 H,O. C = 61-14 ; H = 6.28.C,,H,~0,N3 requires C = 61-1 ; H = 6.2 per cent.3-Methyl-1 -hydkdone, C,H4<co->CH,. CHMeThis substance is formed when ethyl 3-methyl-1-hydrindone-2-mrboxylate, suspended in 20 per cent. sulphuric acid, is treated witha current of steam. The ketone, as it is formed, passes over with thesteam, and can be obtained as a colourless liquid, boiling at 250'1756 mm., on extracting the distillate with ether :0,1693 gave 0.5118 CO, and 0*1080 'H,O.CloH,,O requires C = 82.2 ; H = 6.8 per cent.The phenylhydrazone, C,,H,,N,, separates as an oil, which solidifieson scratching, when a hot solution of the ketone in dilute acetic acidis mixed with a hot solution of phenylhydrazine acetate.Itcrystallises from dilute alcohol in glistening yellow plates, meltingat 95':C = 82.46 ; H = 7-01,0.1684 gave 05025 CO, and 0.1035 .H,O.The semicarbazone, C11H1,0N3, prepared in the usual way,0.1049 gave 0.2493 CO, and 0.0614 H,O.C = 81-27 ; H = 6.69.C,,H,,N, requires C = 81.4 ; H = 6.8 per cent.crystallises from dilute alcohol in slender needles, melting at 190' :C = 64-81 ; H = 6.50.C,,H,,ON, requires C = 65.0 ; H = 6.4 per cent.EthyZ a-o-Dicpano-P-pheny Zpropionate, CN*C,H,*CH,*CH( CN)*CO,Et.-The condensation of o-cyaaobenzyl chloride with the sodiumcompound of ethyl cyanoacetate, if carried out in the usual manner,yields very little of the above product or of ethyl 1-imino-2-cyano-hydrindene, the chief compound formed being ethyl a-oo-tricyano-Pp-diphenylisobutyrate, which is produced to the extent of about 80 percent.of the theoretical quantity. If, however, the condensation iseffected in the following manner, a good yield of the normal productcan be obtained : Two grams of sodium are dissolved in 50 C.C. ofalcohol, and 10.3 grams of ethyl cyanoacetate are added, the solutionbeing kept warm in order to prevent the sodium compound fro2276 MITCHELL AND THORPE: THE FORMATION ANDcaking. The hot sodium compound is then added to a hot solution of13-5 grams of o-cyanobenzyl chloride in 20 C.C. of alcohol, and tbemixture kept hot until the reaction is finished.The bulk of thealcohol is then distilled off, and the residue diluted with water, whenan oil separates which becomes partly solid. On extracting withether, the solid, which was found to be ethyl a-oo-tricyano-PP’-dipheoyl-isobutyrate (see p. ZZSO), remains undissolved, and can be separatedfrom the ethereal solution of the oil by filtration. The residueleft on evaporating the dried ethereal extract is then distilled underdiminished pressure. It is a viscid, colourless liquid, which boils a t220°/20 mm. :0.1921 gave 0,4844 CO, and 0-0934 H,O.C,,B,,O,N, requires C = 68.9 ; H = 5.3 per cent,EtlLyZ a-o-dic?lano-P-pAen?l~p~opionate does not dissolve in coldconcentrated hydrochloric acid. Its structure is shown by theformation of o-carboxy-P-phenylpropionic acid from it on completehydrolysis with dilute snlphuric acid.C = 68-77 ; H --= 5.40.I -1mino-2-c yanohydrindem, C,H,<~&~-~>CH* CN.This substance may be prepared either by the action of sodiumet hoxide on ethyl a-o-dicyano-p-phenylpropionate, or by the directcondensation of o-cyanobenzyl chloride with the sodium compound ofethyl cyanoacetate in the presence of excess of sodium ethoxide.(1) Fyom Ethyl a-o-Bicyano-P-pheny1propionate.-In this preparation10 grams of the dicarboxylic ester are dissolved in 25 C.C.of alcohol,and mixed with 1 C.C. of a solution of sodium ethoxide containing1 gram of sodium dissolved in 10 C.C. of alcohol. The solution, whichbecomes brown in colour, is then warmed on the water-bath for fiveminutes, when it is cooled and diluted with an eqaal volume of water.Crystals separate on scratching, which, when recrystallised frombenzene, form large prisms, melting at 137’.The mother liquor from the condensation, when extracted by ether,yields a residue on evaporating the solvent, which when carefullyfractionated furnishes a considerable amount of ethyl carbonate,boiling at 126’.(Found, C = 50.58 ; H = 8.61. Calc., C= 50.8 ;H =I 8.5 per cent.)(2) Prom Ethyl Sodiocyanoacetate and o-QyanobenxyZ Chlo&de,-This method of preparation always yields the imino-nitrile mixed withsome ethyl a-oo-tricyano-fl~-diphenylisobutyrate, from which it cannotbe readily separated. If the ketone is required, it can be quicklyisolated pure by this means, because it is only necessary to treat themixture with hot dilute hydrochloric acid, extract with ether, andshake out the ketone by means of sodium carbonate solution, in ordeREACTIONS OF JMINO-COMPOUNDS.PART XIV. 2277to obtain the pure product on acidifying the alkaline extract. Owingto the ease with which the irnino-compound is hydrolysed to theketone by hydrochloric acid, it is not advisable to separate themixture by the aid of this reagent. Ultimately the following processwas found to give satisfactory results.The condensation was effected in the same manner as described forethyl a-o-dicyano-p-phenylpropionate (see p. 2275), and when thereaction was finished a slight excess of sodium ethoxide was addedand the heating continued for fifteen minutes.Water was then added,and the solid which separated was collected. It was then rubbed withcold methyl alcohol, filtered from the undissolved di-derivative, andprecipitated from the methyl-alcoholic solution by the addition ofwater.After this process had been repeated twice, the dried product wasrecrystallised from benzene, when i t was obtained in large prisms,melting at 137" :0.1898 gave 0.5346 CO, and 0.0907 H,O. C= 76.83 ; H= 5.31.C,,H,N, requires C = 76.9 ; H = 5.1 per cent.1-lrnino-2-cyanohydrindene is at once soluble in cold concentratedhydrochloric acid. When treated in hot acetic acid solution witha solution of phenylhydrazine acetate, it yields the same hydrazone as2-cyano-1-hydrindone (see p.2278).This compound is readily prepared from 1 -imino-2-cyanohydrindeneby hydrolysis with dilute hydrochloric acid, for which purpose thefollowing conditions were found to give the best results. Ten gramsof the imino-compound were dissolved in concentrated hydrochloricacid, and the clear solution poured into twice its volume of boilingwater, the solution being cooled as quickly as possible after theaddition of the acid. An oil separated, which was extracted by ether,the ethereal extract being shaken with dilute sodium carbonatesolution. The alkaline extract was then acidified, when an oil wasprecipitated which solidified on scratching. When collected andrecrystallised from dilute aloohol, it formed small, colourless needles,melting at 73O :0.1738 gave 0.4845 GO, and 010712 H,O.2-Cyano-1-hydrindone is readily soluble in dilute alkaline carbonatesolutions.When dissolved in dilute aqueous potassium hydroxide, thepotatwium salt separates on the addition of excess of the alkali. Theketone gives a green coloration in alcoholic solution with ferric chloride.C = 76.03 ; H = 4.55.C,,H,ON requires C = 76.4 ; H = 4.5 per cent2278 MITCHELL AND THORPE: THE FORMATION ANDThe phenythydrazone, C,,H,,Ns, is precipitated when a solutionof the ketone in dilute acetic acid is warmed with a solution of phenyl-hydrazine acetate. It separates from alcohol in pale yellow needles,melting at 160° :0.1797 gave 0*5110 CO, and 0.0851 H,O.The 0-bsnxoyl derivative, C17H1102N, can be prepared in small yieldIt crystallises from alcohol inC = 78.16 ; H = 4-10,C = 7’7.55 ; H = 5.26.Cl6H1,N, requires C -- 77.7 ; H = 5.3 per cent.by the Schotten-Baumann method.long, colourless needles, melting a t 101.5O :0,1355 gave 0.3884 CO, and 0.500 H,O.C1’1H1102N requires C =: 78.1 ; H = 4.2 per cent.When 2-cyano-1-hydrindone is boiled with dilute sulphuric acid(10 per cent.) for five hours and the product is distilled in a current ofsteam, a-hydrindone is formed, and passes over with the steam.2-C yano- 3-methox yindene, C,H,<gg2EXX?N.The above methoxy-derivative is the sole product of the methylationof 2-cyano-1-hydrindone whether the alkylation is carried out by meansof sodium methoxide and methyl iodide or whether the pure potassiumsalt is used for the purpose.The preparation can be convenientlyeffected as follows: The potassium salt prepared by the additionof excess of potassium hydroxide to a solution of the ketone in diluteslkali is purified by rubbing with cold ethyl alcohol, and is thensuspended in methyl alcohol and treated with excess of methyl iodide.The mixture is heated on the water-bath until all the salt has passedinto solution, when i t is evaporated on the water-bath and treated withwater. The oil which is then precipitated is extracted with ether, andthe ethereal extract dried and evaporated. The residue distils at185O/20 mm. as a clear, colourless oil, which is quite insoluble in colda1 kali :0.2013 gave 0.5691 CO, and 0.0978 H,O.CllH,ON requires C = 77.2 ; H = 5.3 per cent.The methoxy-structure of the compound is clearly shown by itsbehaviour on hydrolysis, for when it is treated with warm aqueouspotassium hydroxide it slowly dissolves, and when the solution isacidified a crystalline substance melting at 73” is deposited.Thiscompound was shown by direct comparison to be identical with2-cyano- 1 -hydrindone.C = 77.11 ; H = 5.4.Ethyl o-Cyano-a-benxylacetoacetate, CN*C,H,*CH,*CHAc*CO,Et.The conditions employed in preparing this substance were asfollows : 21.7 Grams of ethyl acetoacetate were added to an alcoholiREACTIONS OF IMINO-COMPOUNDS. PART XIV. 2279solution containing 3.5 grams of sodium, and the mixture slowly addedwhile hot t o a hot solution of 25 grams of o-cyanobenzyl chloride in25 C.C.of alcohol. The reaction was vigorous, and when all the sodiumcompound had been added, the product was found t o give a neutralreaction. The greater portion of the alcohol was then distilled off andthe residue mixed with water, when an oil separated, which wasextracted by ether, The product obtained, after evaporating the driedethereal solution, boiled a t 210°/20 mm., forming a viscid, colourlessliquid :0.1972 gave 0-4947 CO, and 0*1100 H20,C,,H,,O,N requires C = 68.6 ; H = 6.1 per cent.Ethyl o-cyano-a-benxylacetoacetate is not affected by cold concentratedhydrochloric acid, and on complete hydrolysis with dilute sulphuricacid yields o-carboxy-P-phenylpropionic acid.C = 68.47 ; H = 6.19.The Transformation of Ethyl o-Cyano-a-benzylacetoacetate into Ethyl1 -Iminohydp.indene-2-carboxy?ate.This conversion was effected by the action of sodium ethoxide in thefollowing manner : Ten grams of the ester were dissolved in alcohol,treated with 1 C.C.of a 10 per cent. solution of sodium ethoxidein alcohol, and then warmed on the water-bath for fifteen minutes.At the end of this time the solution, which had a strong odour ofethyl acetate, was mixed with water, and the solid which thenseparated was filtered. When recrystallised from dilute alcoholit was melted at 98O, and was proved, by direct comparison, to beidentical with ethyl 1-iminohydrindene-2-carboxylate.The aqueous mother liquor from the condensation was saturated withammonium sulphate and extracted with ether, the ethereal extractbeing washed with calcium chloride solution to remove alcohol, dried,and fractionated.The fraction boiling at 78O was collected, and provedfrom its odour and analysis to be ethyl acetate. (Found, C = 54.49 ;H=9.21.When ethyl 1-iminohydrindene-2-carboxylate is required in largequantities, the best method for its preparation is the direct condensa-tion of ethyl sodioacetoacetate and o-cyanobenxyl chloride in thepresence of a slight excess of sodium ethoxide, because in thiscondensation the formation of the di-derivative does not take place toany appreciable extent. The product of the condensation, whichsolidifies on pouring into water, is separated in a pure condition bydissolving in concentrated hydrochloric acid, filtering, and extractingthe filtrate after dilution with water with ether, The yield of ethyl1-iminohydrindene-2-carboxylate under these conditions is 80 per cent.of the theoretical.Calc., C e 5 4 .7 ; €€=9*1 per cent.2280 MITCHELL AND THORPE: THE FORMATION AND?'he Di-o-cyanobenxy? DeYivatiz;es Formed in the ForegoingCondernscctions.Ethyl di-o-cyanobenzylmalonate, (CN'C,H,*CH,),C(CO,Et),, ande t hy 1 di-o-cy a nobenz y lace toace t ate, (CN-C,H,* CH,),CAc* CO,E t, ob-tained as by-products i n the condensations of o-cganobenzyl chloridewith ethyl sodiomalonate and ethyl sodioacetoacetate respectively,have been prepared and described by Gabriel and Hausmann (Zoc. cit.),but they were not further investigated. The following is a descriptionof the products which these compounds yield on hydrolysis, as well asof those derived from ethyl a-oo-tricyano-PP'-diphenylisobutyrate,which has not been prepared before.Ethyl u-oo-Tricyano-pp1-dip~7LenyZisobutyrate,(CN*C,H,*CH,),C( CN)*CO,Et.This substance is obtained as a by-product in the condensationof ethyl sodiocyanoacetate with o-cyanobenzyl chloride, and remuinsundissolved after the imino-nitrile has been separated by the methoddescribed on page 2276.It separates from alcohol in small,colourless needles, melting a t 123' :0.2121 gave 0.5695 CO, and 0.1002 H,O.The ester is sparingly soluble in ether.C = 73-27 ; H = 5.25.C2,Hl7O2N3 requires C = 13.5 ; H = 5.0 per cent.a-oo-Tricyano-pP'-diphenylisobutyric Acid,(CN*C,H,*CH,),C(CN)*CO,H.When ethyl a-oo-tricyano-@'-diphenylisobutyrate is warmed withaqueous potassium hydroxide, it passes into solution, and if, when allhas dissolved, the solution is cooled, an oily potassium salt separates,which dissolves on the addition of more water.On acidification, theclear solution deposits an oil which solidifies on stirring, and the solidcan then be recrystallised from hot alcohol, from which solvent itseparates as a microcrystalline powder melting at 175' with vigorousevolution of gas :0.1508 gave 0.3988 CO, and 0.0569 H,O. C = 72.12 ; H = 4.18.C,,H,,O,N, requires C = 73.4 ; H = 4.1 per cent.oo-Dicyano-PP'-d~phenyZiso6utyronitr~Ze, (CN*C6H,*CH2),CH*CN.This substance is formed when the above carboxylic a J d is heatedat 170' in a bath of sulphuric acid until all carbon dioxide has beenevolved. The dark-coloured residue solidifies on being rubbed witREACTIONS OF IMINO-COMPOUNDS. PART XIV.2281ether, and at the same time becomes colourless. It may be furtherpurified by recrystallisation from methyl alcohol, from which solvent itseparates in large, flattened needles, melting at 132-133" :N = 25.3. 0.1733 gave 23.2 C.C. N2 at 1 8 O and 733 mm.The trinitrile is readily soluble in hot methyl or ethyl alcohol andCl8Rl3N3 requires N = 15.5 per cent.It is sparingly soluble in ether or light petroleum. in hot benzene.oo- D icarboxy-PP-diphenyliso but yric Acid,(CO,H*C,H,* CH,),CH*CO,H.This acid is formed by the ultimate hydrolysis of the di-o-cyanobenzylderivatives described above.From Ethyl a-oo-T&yano-~~'-diphenyZisobuty~*ate.-!I!he ester isdissolved in concentrated sulphuric acid, and water is added until thesolution just remains clear, when i t is boiled on the sand-bath until anoil begins to separate, More water is then added until the solutionis clear, when it is again heated until oil begins to separate.Thisprocess is continued for two hours, when the solution, which on coolingdeposits large quantitias of oil, is extracted by ether. When theethereal extract is shaken with aqueous sodium carbonate and thealkaline extract acidified, a gummy acid is precipitated, which readilysolidifies when rubbed with methyl alcohol, and crystallises fromdilute alcohol in small needles, melting at 210'.Rrom Ethyl Di-o-cyanobenxyZmaZo~ate.-When this ester is boiled withan alcoholic solution containing one and a-half times the quantity ofpotassium hydroxide calculated for complete hydrolysis, and the heat-ing is continued until all ammonia has apparently been given off, thesolution on evaporating and acidifying yields a resinous acid. Whenthis resin is extracted by ether, a large quantity of crystalline solidremains undissolved by the ether, and can be isolated by filtration.It can be recrystallised from much hot water, and is then obtained insmall needles, melting at 2 2 7 O . The analysis and properties of thissubstance showed it to be the acid diamide of the formula(NH2*CO*C6H4-CH2),CH*C0,H :0.1820 gave 0.4430 CO, and 0.0942 H,O.C,,H,,O,N, requires C = 66.3 ; H = 5.5 per cent.The following titration was also made : 0.2600 required 8.1 C.C. ofN/lO-NaOH for neutralisation, whereas this amount of a monobasicacid, C,813[,804N2, requires 8.00 C.C.The acid diamide when boiled for some time with excess of aqueouspotassium hydroxide evolves ammonia, and the solution on rlcidifyingyields a gummy acid, which soldifies on rubbing with alcohol, Whenrecrystallised from dilute alcohol it forms small needles, melting atVOL. XCVII. 7 KC=66*38; H=5*762282 THORPE AND SIMMONDS: LEAD SILICATES IH210°, which are identical with those of the acid prepared from ethyla-oo-tricyano-P~-diyhenylisobutyrate :0.2735 gave 0,4190 CO, and 0,0777 H,O.C18Hls06 requires C = 65.9 ; H = 4.9 per cent.The ethereal solution after the separation of the above acid diamideyields an oily residue on evaporation, which solidifies when rubbedwith alcohol, It crystallises from dilute alcohol in yellow prisms,which melt a t 142O, and yields a semicarbazone melting a t 256'. Wehave not as yet succeeded i n assigning any satisfactory formulze tothese substances.C = 65.86 ; H = 4.97.Some of the expense entailed by this research has been met bygrants from the Government Grant Committee of the Royal Society,for which we desire to express our indebtedness.THE SORBY RESEARCH LABORATORY,THE UNIVERSITY, SHEFFIELD