NITRO- ARYLAZO- AND AMINO-BLYOXALINES. 21 7 XXVI .-Xitro- Arylazo- and Amino-glyoxalines. By ROBERT GEORGE FARGHER and FKANK LEE PYMAN. THIS investigation was begun with the object of effecting the synthesis of purine derivatives by a method complementary to those which have been employed hitherto. In these the pyrimidine nucleus is first built up and the glyoxaline ring closed subsequently. We proposed to prepare 4-aminoglyoxaline-5-carb-oxylic' acid,* condense it with cyanic acid and eliminate water with t,he production of xanthine. Such a synthesis would be of interest in view of the suggestion that purine derivatives originate from histidine in the animal body (compare Hopkins T. 1916 109 629). Although the starting material for the proposed synthesis, 4-aminoglyoxaline-5-carboxylic acid was unknown we did not anticipate that its preparation woald offer any serious difficulty.We have however so far failed to obtain this substance and now give an account of our attempts to prepare this and other amino-substituted glyoxdines. An account of the investigation may be subdivided under three headings first the preparation of the glyoxalines and their carb-oxylic acids which were required as starting materials ; second, the preparation and properties of nitroglyoxalines; and last the preparation and properties of arylazoglyoxalines. (I) The I'reparatiom o,f Glyoxalines and their Carboxylic Acids. -For the purpose of this investigation it was necessary to prepare considerable quantities of glyoxaline-4 5-dicarboxylic acid the most convenient source of glyoxaline.This acid was first prepared by Maquenne (Ann. chim. phys. 1891 [vi] 24 525> by mixing aqueous solutions of nitrotartaric acid and hexamethylenetetramine, adding ainmoiiia and allofwing the mixture t o become hot and sub-sequently by Dedichen ( H e r . 1906 39 1835) who replaced the ~~er;amethylenetetramilze by formaldehyde. We have carried out a large number of experiments on the best conditions for the pre-* In glyoxalines containing a free imino-group the 4- and 5-positions are equivalent 21s >'ARCHER AND PYMAN: paration of this acid and find that to obtain a good yield it. is essential that the reaction mixture should be kept cold. A number of experiments were carried out? with the object of effecting t.he partial decarboxylation of glyoxaline-4 5-dicarboxylic acid and thus producing glyoxaline-4-carboxylic acid by a more convenient and economical process than that previously employed, where six operations are required in its synthesis from citric acid through 4-hydroxymethylglyoxaline (T.1911 99 668 ; 1916 109, 186). When the acid is heated with water 10 per cent. hydro-chloric acid or concentrated hydrochloric acid little decarboxyla-tion takes place below 180° but above this temperature the action proceeds more readily glyoxaline being the main product whilst a small proportion of glyoxaline-4-carboxylic acid can be isolated provided that the heating has not been too prolonged. m7hen the acid is heated with an excess of concentrated ammonia a t 180° to 200° the main product is glyoxaline," and a similar result is obtained by heating the aqueous solution of the mono-sodium salt.The desired result can be obtained however by boiling the acid with aniline when the aniEide of glyoxaline-4-carboxylic acid is formed in a yield amounting to 45 per cent. of the theoretical. From this the acid is readily prepared by hydrolysis. For the purpose of orientation it was necessary to prepare glyoxalinea substituted in the 2- 4 5- and 2 4 5-positions. The 2-alkylglyoxalines were prepared by suitable modifications of Maquenne's methods. From 2-methylglyoxaline-4 5-dicarboxylic acid 2-met?~ylglyo~aline-4-cnrb ozylic acid was obtained through its nnilide. As representatives of 4 5- and 2 4 5-substituted glyoxalines, 4 5-dimethylglyoxaline and 2 4 5-trimethylglyoxaline were pre-pared by modification of known methods.(2) Nitro~ZyoxaZines.-The nitration of various glyoxalines has led to the formation of mononitroglyoxalines in the hands of several observers. I n some casw the nitro-group evidently enters the 4-(or 5-)position since no other position is vacant; for instance, in the nitration of 2-methylthiol-1-phenyl(and 1-methy1)glyoxaline * We were unable to find any evidence of the formation of the imide of glyoxaline-4 5-dicarboxylic acid from which the desired 5-aminoglyoxaline-4-carboxylic acid might have been obtained by the action of hypobromous acid NITRO- ARYLAZO- AND AMINO-GLYOXALINXS. 219 (I) (Wohl and Marckwald Ber. 1888 22 568 1353) and 2:4di-methylglyoxaline (11) (Windaus Ber.1909 42 758) : >CMe. gH*NH CMe-N (1.1 (11.) The orientation of the nitro-group in nitroglgoxaline itself (Rung and M. Behrend tl?~nalen 1892 271 28; R. Behrend and Schmitz ibid. 1893 277 338) and in nitro-4-methyiglyoxaline has not been determined previously but an indication that the latter contains the nitro-group in the 5-position is afforded by Windaus's observation (ZOC. cit.) of its close similarity to 5-nitm-2 4-dimethyl-gl yoxaline. Moreover whilst) 412it7.0-2-rnethyl~ly~xal~ne is readily prepared, we were unable to obtain a nitro-derivative of 4:5-dimethyl-glyoxaline for in this case part of the base was completely oxidised, whilst a considerable proportion remained unchanged and the only isolable derivative was the nitrate of 4-methylglyoxaline-5-carb-oxylic acid which has been described by Gerngross (Ber.1912, 45 509). The inability of a glyoxaline substituted in both the 4- and 5-positions to form a nitro-derivative indicates that the nitro-glyoxalines contain the substituent in the &(or 5-)position. This view is confirmed by their behaviour on reduction. Wohl and Marckwald (Zoc. cit.) attempted to reduce the 4-(or 5-)nitro-2-methylthiol-l-phenyl-(and l-methy1)glyoxalines to the correspond-ing amines but obtained only decomposition products including methyl mercaptan. Similarly we find that 4-nitro-%methyl-glyoxaline undergoes fission on reduction with tin and hydrochloric acid two of the three atoms of nitrogen in the molecule appearing in the form of ammonia.* Since precisely the same result is obtained with nitroglyoxaline and nitro-4-methylglyoxaline whilst it is shown below that 2-aminoglyoxalines are stable it is clear that these nitro-derivatives are 4-nitroglyoxaline and 5-nitro-4-methyl-glyoxaline respectively.Before we had arrived at this conclusion we were anxious to prepare some of the nitroglyoxaline-4-carboxylic acid which Windaus and Opitz (Ber. 1911 44 1721) obtained by the action of boiling 25 per cent. nitric acid on 4-P-hydroxyethylglyoxaline. ' The first stage in the disintegration of the 4-aminoglyoxalines is probably the elimination of the amino-group as ammonia with the formation of a glyoxalone for certain members of the purine group-also derivatives of 4-aminoglyoxaline-have been shown to undergo hydrolysis in this mmer (compare for instance Tafel and Mayer Bey.1908 41 2546; Biltz Ber., 1910,43 15S9) 220 FARGHER AND PYMAN: These authors state that the yield of $-~-hydroxyethylglyoxaline, obtained by the action of barium nitrite on 4-B-aminoethyl-glyoxaline hydrochloride was so poor that the nitro-compound was not available in sufficient quantity for further study. It appeared to us however that this nitro-compound might be obtaiaed by t’he action of nitric acid on other more readily accessible derivatives of glyoxaline containing a side-chain of carbon atoms in the 4-posi-tion and in the first place we employed compounds containing two carbon atoms in the side-chains like Windaus’s starting material. The results were disappointigg ; 4-P-aminoethyl-glyoxaline when boiled with 50 per cent.nitric acid for nine hours was mainly recovered unchanged whilst 4-cyanomethyl-glyoxaline was converted under the same conditions almost quanti-tatively into glyoxaline-4-acetic acid. Atltempt.s t o nitrate glyoxaline-4-carboxylic acid and glyoxaline-4 5-dica,rboxylic acid were likewise unsuccessful. The prospect of nitrating 4-hydroxy-methylglyoxaline was not hopeful for it has bem shown previously (T. 1916 109 186) that hot concentrated nitric acid converts it into glyoxaline-4-f ormaldehyde and glyoxaline-4-carboxylic acid. It has now been found that the alcohol gives the same products when digested oil the water-bath with fuming nitric acid whilst it can be recovered almost quantitatively after boiling with ten parts of 25 per cent.nitric acid for four hours. On the other hand the nitration of 4-hydroxymethylglyoxaline with nitric and sulphuric acids gave rise to a product which was not obtained in crystalline form but further study of this was olmit,ted in view of the peculiar behaviour of the simple nitroglyoxalines on reduction. (3) A ryluzoglyoxa1ines.-The constitution of the arylazo-deriv-atives of simple glyoxalines has not been settled hitherto. Rung and Behrend (Amnalen 1892 271 ZS) who first isolated benzene-azoglyoxaline considered it to be a diazoimino-compound (I), because boiling acids decomposed it with the formation of nitrogen > CH, EH*N(N:NPh) CH-N’ and glyoxaline. Burian (Ber. 1904 37 696) wno prepared many arylazoglyoxalines from diazobenzenep-sulphonic acid and various glyoxalines adopted the same view of the constitution of these compounds on other grounds namely because all the glyoxaiines substituted in some or all of the 2- 4- and 5-positions which he examined coupled with the diazonium salt whilst 1 -substituted glyoxalines did not.Pauly (Zeitsch. yhysiol. Ghem. 1904 42, SOS) however pointed out the possibility that the arylazo NITRO- ARYLAZO- AND AMINO-GLYOXALINES. 221 glyoxalines were true Cl-azecompounds (11) similar to those obtained from pyrrole and later (ibid. 1915 94 284) attributed the probable formula (111) given below to the compound obtained by the action of diazotised arsanilic acid on histidine owing to its stability towards acids. Whilst in the case of hhese simple glyoxalines the orientation of the arylazo-group is uncertain the constitution of the arylazopurines is known for Hans Fischer (Zeitsch.physiol. Chem. 1909 60 69) has shown that the arylazo-group enters the 8-position of the purine nucleus-the 2-position of its glyoxaline ring-by reducing arylazopurines t o 8-aminopurines. I n view of this result it appeared to us probable that the arylazclderivatives of simple glyoxalines were also C-azo-compounds, as Pauly suggested and it was of interest to determine whether the arylazo-group entered the 2- or the 4-position and the nature of the products obtained on reduction. The benzeneazoglyoxaline of Rung and Behrend was first examined. By the method of these workers it is obtained in poor yield but by the action of benzenediazoniuin chloride on one mole-cular proportion of glyoxaline in an excess of aqueous sodium carbonate it is readily obtained mixed with a little 2 4 5-tm's-b enzeneazoglyoxdine.2-Benzeneazoglyoxaline melts a t 190° (con-.) and it' is therefore evident that the specimen prepared by Rung and Behrend melting a t 177-17807 was impure. The pure substance is reasonably stable towards boiling 10 per cent. hydro-chloric acid for a considerable proportion can be recoverdd un-changed after two hours. The constitution of 2-benzeneazoglyoxaline (IV) was proved by reduction. With zinc dust1 and hot acetic acid it yields aniline and glycocyamidine%* (V) the formation of the latter showing that the benzeneazo-group is attacheld to the 2-position of the ring.>C:NH + PhNH YH,*NH EH*NH CH-N CO-N H >C*N:NPh + (IV.) (V.1 * The conversion of glyoxaline into glycocyamidine and 2-Elminoglyox&ne, both derivatives of guanidine is of biochemical interest firstly because creatinine is the N-methyl derivative of glycocyamidine and secondly on account of the similar behaviour of histidine and arginine in purine metabolism (oompare Hopkins koc.dt.) 222 FARGRER AND PYMAN : This result is confirmed by the formation of a small amount of guanidine on the reduction of 2-benzeneazoglyoxaline with stannous chloride. In this reduction a small quantity of 2-aminogZyoxaline is formed and some aniline but the main product is 2-amino-4-p-amircophenylglyoxaline (VI) a compound resulting from a change of the benzidine type.It is also formed in small proportion in the reduction with zinc dust and acetic acid. (VIII .) Its constitution was proved by the oxidation of its diacetyl deriv-ative with potassium permanganate when p-acetylaminobenzoic acid was formed. This result eliminated the possibility that the compound had one of the two formulze (VII) or (VIII) represent-ing substances formed by a change of the semidine type. The occurrence of a rearrangement of the benzidine type in a five-membered heterocyclic nucleus seems reinarkable at first sight but a cIoser inspection of the formula shows that the conjugated system connecting the 2- and 5-carbon atoms of the glyoxaline ring is similar t-o that existing in the benzene nucleus. NH- NH 6 Is H C*NHS /\ s/\ /\ N/\ 1 I CHI >NH - I 1 1 CHI P H .\/ C \/ I-_-_ \/ CH \P Whilst no other case of the benzidine type od change in a heterocyclic nucleus has been observed previously so far as we are aware Michaelis and Schafer (Annnlcn 1915 407 229) have obtained by the reduction of l-phenyl-3-methyl-4-benzeneazo-pyrazole (IX) the two isomerides (X) and (XI) which result from the two possible changes of the semidine type. N PI1 K P h NPh (IX.) ( X . 1 (XI.) Owing to the formation of 2-amino-4-paminophenyiglyoxaline in the reduction of 2-benzeneazoglyoxaline the yield of Z-amino NITRO- ARYLAZO- AND AMINO-GLYOXALINES. 223 glyoxaline is small so for the preparation of this substance the reduction of an arylazoglyoxaline containing a substituent in the para-position of the benzene nucleus was undertaken.2-p-Bromobenzeneazoglyoxdine is the main product of the inter-action of p-bromobenzenediazonium chloride and glyoxaline in aqueous sodium carbonate only a very small proportion of 4-p-bromobenzeneazogl?/o.xaZine being formed. The reduction of 2-p-bromobenzeneazoglyoxaline with stannous chloride gave 2-aminoglyoxaline in a yield of 56 per cent. of the theoretical, together with aniline guanidine some ?-amino-4-paminophenyl-gIyoxaline and a small quantity of a base C,R,N,Br which is probably 2-Eif -bromo-2~-am~noan~linoylyoxaLine (compare p. 246). 2-Aminoglyoxaline is a monacidic base yielding crystalline salts, but the free base has not been obtained in a crystalline form. 8H'NH>C*NH, CH-N (XII.) (XITI.) For this compound the tautomeric formulz (XII) and (XIII) are possible.The first is supported by the production of a red colour when the substance is mixed with sodium diazobenzene-p-sulphonate and by the fact that after treatment with nitrous acid it couples with phenols. An indication that it can also react according to the formula (XII1)-which represents an unsaturated compound no longer containing the glyoxaline ring-is given by its behaviour towards permanganate for 2-aminoglyoxaline and all t.he substituted 2-aminoglyoxalines described in this paper reduce cold aqueous acid potassium permanganate and in this respect resemble the 2-thiolglyoxalines (compare T. 1911,99 2173), whereas glyoxaline and its homologues are stable t o this reagent, although they reduce alkaline permanganate giving green solutions.2-Aminoglyoxaline yields a mononcet?p? and a monobenzoyl deriv-ative which are stable to cold aqueous acid permanganate. 2-Aminoglyoxaline does n o t combine with benzaldehyde in acetic acid solution. Moreover 2 - am in 0-5 - p-amino p h en y l-4-m ethyl-glyoxaline (XVII) yields only a monoberizylidene compound under these conditions doubtless 2-amino-5-~~-benzylideneaminophenyl-4-methylglyoxaline. This behavio'ur therefore serves to differentiate between 2-aminoglyoxalines and homologues of aniline and is employed later in the determination 01 constitution. The action of benzenediazonium chloride on 4-methylglyoxaline proceeded quite differently from its action on glyoxaline. Instead of the 2-subst,itutsd arylazo-compound being formed predominantly 224 FARGHER AND PYMAN: nearly equal quantities of 2- benzeneazo-4-methylglyozali~e, 5-benzeneazo-4rnethylglyoxaZine and 2 ; 5- bisbenzeneazo-4-methyl-glyoxaline were obtained.The constitution of 2-benzeneazo-4-methylglyoxaline (XIV) follows from the fact that it yields 011 reduction with zinc dust and acetic acid alacreatinine (XV) a compound previously synthesised by Baumann ( A nnalen 1873, 167 83) by the elimination of water from acid (XVI). >C*N:NPh + EMe- N TT CH--N->C:NH f-$lHMe*NH CO-NH (XV. 1 This change is precisely similar to the a-guanidinopropionic pMe*NH>C:NH C0,H NH, (XVI.) formation of glyco-cyamidine from 2-benzeneazoglyoxaline. 2-Benzeneazo-4-methyl-glyoxaline behaves in the same way as 2-benzeneazoglyoxaline on reduction with stannous chloride the principal product of the reaction 2 - amino - 5 - p - nminophenyl - 4 - nzethylglyoxaline (XVII) a compound having similar properties to 2-amino-4-p-aminophenylgl yoxaline.being (XVII.) (XVIII. ) The constitution of 5-benzeneazo-4-methylglyoxaline (XVIII) could not be proved directly as in the case of the 2-isomeride. On reduction aniline and a considerable amount of ammonia were formed,' together with other products which included a base, C,Hl,0N2 (p. 254) when stannous chloride was employed as the reducing agentl and a base C,,H,,ON (p. 255) when zinc dust and acetic acid were used. The disintegration of the molecule indicated by the formation of ammonia is similar to that occurring in the reduction of the 4-nitroglyoxa!ines and affords evidence that the constitution of the compound is represented correctly by the formula of 5-benzeneazo-4-methylglyoxaline.The formula is sup-ported by the fact' that the compound is soluble in dilute aqueous sodium hydroxide which indicates that' the imino-group is unsub-stituted. Moreover it is fairly stable towards boiling dilute acids. Its properties are not therefore in accord with those of a compound represented by the alternative formula 1-benzeneazo4-methyl-gl yoxaline. That aryldiazonium salis are capable of substituting the 4-posi-tion of the glyoxaline ring follows from the reduction of 2-phenyl NITRO- ARYLAZO- AND AMINO-OLYOXALINES. 225 4-p-bromobemzeneazv$yoxaZine C,,H,,N,Br for a compound, C,,HI3N,Br is produced which is evidently derived from the corre-sponding hydrazo-compoand by a change of the semidine or benz-idine type (compare p.257). The polyarylazoglyoxalines-2 4 5-trisbenzeneazoglyoxaline and 2 5-bisbenzeneazo-4-met~hylglyoxaline-are insoluble in dilute mineral acids and are decomposed on boiling with 10 per cent. hydrochloric acid. Nevertheless we regard them as C-azo-com-pounds because they are soluble to some extent in aqueous sodium hydroxide I n the case of the second,compound we have estab-lished the fact that it is precipitated unchanged from its solut,ion in aqueous sodium hydroxide by means of acetic acid. The fact that the number of arylazo-groups in the polyarylazo-compounds corresponds with the number of nuclear methine groups ia the parent glyoxaline points in the same direction.The interaction of glyoxaline-4 5-dicarboxylic acid and diazo-benzene-p-sulphonic acid was studied by Burian (Zoc. cit.) who found that carbon dioxide was liberated and described a product forming yellow needles or red microscopic prisms which gave on analysis results indicating that it was a compound derived from one niolecular proportion of diazobenzenep-sulphonic acid and one of glyoxaline - 4 - carboixylic acid S03H*C6H4-N:N=C3H,N2*C0,H. Burian regarded this as a l-substituted arylazoglyoxaline butl we thought it more probable that t.he arylazo-group had displaced a carboxyl group in the 4-(or 5-)position and that' the compound was 5 -p-sulphobenzeneaz~glyoxaline-4-carl1oxylic.acid (XIX). (XIX.) This compound would yield 5-aminoglyoxaline-4-carboxylic acid if a suitable method of reduction could be found and we there-fore attempted to repeat its preparation but were unable to do so. We can confirm Burian's statement that carbon dioxide is liberated in the reaction but find the yield of this to be only about 40 per cent. of tihe theoretical much less than he states. More-over we have isolated in a yield of about 30 per cent. of the theoretical the condensation product of diazobenzene-psulphonic acid and glyoxaline-4 5-dicarboxylic acid namely 2-psdpho-benzeneazoglyoxdine-4 5-dicarboxylic acid ( X X ) . No other definite compound could be isolated from the reactlion mixture and it appears to us that the colmpound dwcribed by Burian was prob-ably a mixture of our acid with its sodium salt.2-pSulphobenzeneaz~glyoxaline-4 5-dicarboxylic acid yields o 226 FARGHER AND PYMAN: reduction with sodium hyposulphite sulphanilic acid and 2-amino-glyoxaline-4 ; 5-dicarbozylic acid (XXI). With the object of eliminating the elements of carbon dioxide, this acid was heated with water for twelve hours a t 170° whell carbon dioxide and approximately one molecular proportion o i' ammonia were liberated but no other fission product could be identified. On the other hand when boiled with aniline for six hours it gave a quantity of 2-aminoglyoxaline. Whilst we were uiiable to isolate 5-11-sulphobenzcneazo-glyoxaline-4-carboxylic acid (XIX) from the products of the inter-action of diazobenzene-psulphonic acid and glyoxaline-4 5 clicarb-oxylic acid the liberation of carbon dioxide indicates that sub-stlitution in the 4-(or 5-)position takes place t o some extent.More-over we can confirm the fact that 2-methylglyoxa!ine-4 5-dicarb-oxylic acid couples with sodium diazobenzene-psulphonate in aqueous sodium carbonate. On the other hand 2 4 5-trimethylglyoxaline (XXII) which contains a free imino-group but no other hydrogen atom attached >CMe fiMe*NH CMe-N (XXII.) to the nucleus does not couple with sodium diazobenzene-p-sulphonate. Further a striking difference is exhibited between the facilities with which 2-amino-4-p-aminophenylglyoxaline (VI) containing a displaceable hydrogen atom in the glyoxaline nucleus and its methyl homologue (XVII) reach with sodium diazobenzene-psulphonate in aqueous sodium carbonate.The first gives the characteristic intense cherry-red colour immediately whilst the second -gives a pale orange cdour which deepens on keeping and is probably due to the participation of the aminophenyl group. On reviewing these results and those of previous investigators it appears t o us that glyoxalines in order t o be capable of coupling, must contain a free iminwgroup and also a hydrogen atom or some other displaceable group such as the carboxyl group in one of the 2- 4- or 5-positions and that the arylazoglyoxalines hitherto pre-pared are C-azo-compounds. The litmeratwe of the arylazoglyoxalines contains one possible exception to this generalisation-the compound (orange needles melting a t 120-122°) described by Burian (lo$.cit.) as having been obtained by the action of diazotised benzidine on 2-thiol-4 5 NITRO - ARYLAZO- AND AMINO-GLY 0 XALXNES. 2 2 7 diphenylglyoxaline. Since we found that 2-thiol-4 5-dimethyl-glyoxaline and 2-thiol4 5-diphenylglyoxaline only gave pale orange colorations with sodium diazobenzene-p-sulphonate we repeated Burian's preparation. We failed however to confirm his results, but isolated from the product as main constituents much un-changed 2-thiol-4 5-diphenylglyoxaline and a reddish-brown, amorphous compound melting and decomposing above ZOOo which from its low nitrogen content (5.5 per cent.) and the ratio of nitrogen to sulphur (2 1) could not. have been an arylazo-derivative derived from 2-thiol-4 5-diphenylglyoxaline.EXPERIMENTAL. Part I GlyoxcLli?Les a n d t h e i r C a r b o x y l i c A c i d s . l'reparutiuu of G'lyoxali~~e-4 5-clica~boxylic Acid. Twenty-five grams of finely powdered tartaric acid are dissolved in 108 C.C. of nitric acid (D 1*5) and 125 C.C. o€ sulphuric acid are added. The mixture which attains a temperature of about 40° soon begins to deposit crystals and is kept for three to four hours in a cool place. The nitrotartaric acid is collected washed with 50 per cent'. sulphuric acid drained on porous porcelain and stirred immediately with 150 grams of powdered ice until dissolved, when the temperature falls to -5'. The liquid is immersed in a freezing mixture and 100 C.C. of aqueous ammonia (D ,0*880) are added gradually the temperature being kept below Oo.Then 50 C.C. of 40 per cent. aqueous form-aldehyde are added slowly keeping the temperature below loo. The product is removed from the freezing mixture after three to four hours and kept overnight. It is then mixed with a little alcohol and acidified with hydrochloric acid when 15.5 t o 16.0 grams of glyoxaline-4 5-dicarboxylic acid separate that is about 60 per cent. of the theoretical yield calculated on the quantity of tartaric acid employed. Glyoxaline-4 5-dicarboxylic acid melts and effervesces a t 2880 (corr.). It is soluble in about SO0 parts of boiling water and in aboat 2000 parts of cold water. It is practically in-soluble in the usual organic solvents but dissolves sparingly in pyridine. It is soluble in cotncentrated mineral acids but is precipitated unchanged on dilution with water.The moao-aodzz~m salt which crystallism from water as a felted mass Qf feathery needles containing lH,O (Found H20 = 9.4 ; in dried salt Na=12.S. @ale. H20=9'2; Na=12'9 per cent.) is sparingly soluble in water but readily so in aqueous sodium hydr 228 FARGHER AND =MAN: oxide probably owing to %he formation of a disodiitm salt in solu-tion. Moreover the addition of alcohol to a solution of the acid in sufficient aqueous sodium hydroxide; to form the disodium salt causes the precipitation of a granular deposit approximating in composition to the disodium salt. (Found in salt dried at l l O o , Na = 21.2. C,H,O,N,Na requires Na = 23.0 per cent.) The acid is very stable towards nitric acid; after boiling it with ten times its weight of concentrated nitric acid for twenty-four hours more than 90 per cent.was recovered unchanged whilst similar results were obtained in a sealed tube a t 130° and when the acid was boiled with equal parts of nitric and sulphuric acids. The acid is very resistant to esterification for after boiling with alcoholic sulphuric acid for twenty-four hours 95 per cent. was recovered unchanged. ?'he Preparation of Glyoxaline . One hundred grams of glyoxaline-4 5-dicai%oxylic acid were distilled under normal pressure in quantities of 4 grams from a small flask into a long wide air condenser. The distillate which had solidified in the condenser was crystaliised from benzene and gave a 92 per cent. yield of the pure base. GtyoxaZirzc piciute crystallises from water in long fine yellow needles which become orange on drying a t loo@ and then melt a t 2120 (corr.) after sintering from 208O.It contains rather more than 1H,O (Found loss a t 10Oo=7.2; in substance dried a t looo, N= 23.3. GZyoxmTTine hydrogen twtrate crystallises from wat'er in fine prisms of Characteristic trapezoidal shape which are anhydrous and melt a t 202O (corr.). It is readily soluble in cold water and is best crystallised from 50 per cent. alcohol (Found N=12.8. c3H,N2,C,H,0 [218*1] requires N = 12.8 per cent.). Qlyoxaline hydrogen oxalate crystallises from water as a felted mass of prismatic needles which are anhydrous and melt a t 232O (corr.) after sintering from 230O. It is soluble in five or six parts of boiling water but much less so in cold water (Found N= 17.8.Calc. N=17*7 per cent.). C,H,N,,C,H,O,N [297*1] requires N=23.6 per cent,). Action of Boiling Aniline om Glyosalinel 5-dicurboxylic Acid: Fomzatkom of Glyoxaline-4-curb oxyanilide and Glyoxaline. Five grams of glyoxaline-4 5-dicarboxylic acid were h i l e d with 50 C.C. of aniline for nine hours under a reflux condenser whe NITRO- ARYLAZO- AND AMINO-GLYOXALINES. 229 the acid gradually passed into solution. The product was mixed with water and subjected to distillation with steam until the excem of aniline had been removed. The residual aqueous solution was filtered from a small quantity of resinous matter whilst still hot, when the filtrate a t once began t o deposit the anilide as a felted mass of fine needles.The first crop amounted to 2.6 grams and a further quantity of 0.1 gram waa obtained on concentrating the mother liquor. The filtrate from this gave on acidification 0.1 gram of glyoxaline-4 5-dicarboxylic acid but no glyoxaline-4-carboxylic acid was found. The final mother liquor when mixed with sodium carbonate evaporated to dryness and extracted with benzene gave 0.9 gram of glyoxaline. Glyoxaline-4-curb oxyanilide crystallises from boiling water in fine colourless needles which are anhydrous and melt a t 227-228O (corr.). It is fairly readily soluble in alcohol but only sparingly so in boiling water and the other usual organic solvents. C,,H,ON (187.15) requires C = 64.2 ; H = 4.9 ; N = 22.5 per cent. Hydrolysis of the A nilide.-The anilide is only slowly hydrolysed by 10 per cent.hydrochloric acid a t looQ but more readily at 1 3 0 O . One gram of the anilide was heated with 10 C.C. of 10 per cent. hydrochloric acid a t 130° for three hours. The resulting solution was evaporated to dryness to remove the excess of acid the residue dissolved in water basified with sodium carbonate and extracted with ether to remove aniline. Sufficient hydrochloric acid was added to render the solution faintly acid to methyl-orange when crystallisation set in almost immediately and 0.42 gram of glyoxaline-4-carboxylic acid was isolated The properties of the acid and its hydrochloride nitrate and picrate agreed with those previously given (T. 191'6 109 199) for the acid prepared by the oxidation of 4-hydroxymethylglyoxaline and the melting points of mixtures of the compounds from the two sources were not depressed.Fou11d C=64*2; H=5.1; N=22.6. 2-Methylglyoxdine-4 5-dicarboxylic Acid. This acid was prepared in an analogous manner to its lower homologue employing a solution of 15 C.C. of freshly distilled acetaldehyde dissolved in 50 C.C. of ice-water in the place of the aqueous formaldehyde. The yield of 2-methylglyoxaline-4 5-dicarboxylic acid containing 1H20 obtained from 25 grams of tartaric acid was 22 grams that is 67 per cent. of the theoretical. VOL cxv. Maquenne (Zoc. tit.) obtained 50 grams of the product fram 100 grams of tartaric acid that is 38 per cent. of the theoretical. Genetally the properties of this acid &re very similar to those of glyoxaline-4 5-dimrheyl-ic acid and it behaves similarly on acid and alkaline hydrolysis.With sodium diazobenzen~~-sulphonate in aqueous sodium carbonate it gives a faint red colour which deepens on keeping, whilst glyoxahe-4 5-dicarboxylic acid gives a deeper red d o u r in the frrst instance. Action of Boiling Aniline o n 2-Methylglyodine-4 Ei-dicarboxylkc A d . Twenty grams of hydrated 2-methylglyoxaline-4 5-dicarboxylic acid when treated with boiling aniline under the same conditions as its lower homologue gave 11 grams of the hydrated anilide of 2-methylglyoxaline-4-carboxylic acid and 3.8 grams of 2-methyl-gl y oxdine. 2-Methylglyoxaline-4-car b oxyandide crystallises from boiling water as a felted mass of cdourless silky needles which contain rather less than 1H,O.It is sparingly soluble in boiling water, but readily so in alcohol. After drying a t l l O o it melte a t 208O (dorr.) . Found loss a t l l O o in three samples=6.9 7.0 7.2. Found in substance dried at l l O o C=65.1 65.6; H=5'7 5.6; C,,Hl10N3,H,0 requires H,O = 8.2 per cent. N=20.9. C1,Hl1ON (201.2) requires C = 65.6 ; H = 5-5 ; N = 20.9 per cent. 2-Methhylgtyoxaline-4-carboxylic acid is obtained in nearly the theorst$cal yield by the hydrolysis of its anilide under similar con-ditions to those already described for glyoxaline4-carboxylic acid. When placed in a bath a t 250° it! melh and effervesces a t 262O (cam.). It crystallises from water in clusters of prismatic needles containing 1H,O. It is soluble in about twenty parts sf boiling water but is practically insoluble in the usual organic solvents.Found loss a t 110°=12*9. Found in %he substance dried a t l l O o C=47-3; H=4-8; C,H60,N2,H,0 requires H,O = 12.5 per cent. N = 219. C,H602N (226.1) requires C=47-6; H=4*8; N=22*2 per cent. With sodium diazobenzenep-sulphonate it gives a red colmr in The hyd?wochloride crystallises from wahr in which it is readily sodium carbonate solution soluble in minute flattened rhombic prisms which are anhydrous. It melts and effervesces a t 2 6 8 O (corn.). Found N=16*9; c1=21.5. The nitrute crystallises from water in which it is very readily soluble in minute rhombic prisms which melt and effervesce a t 190° (corr.) resolidify and on further heating gradually darken, melting a t about 240O.C,H602N2,HC1 (162.6) requires N = 17.2 ; c1= 21.8 per cent. Found C=31*7; H=4*1. The picrute crystallises from water in minute cubes containing 2H,O which is lost a t looo (Found H,O=9*4. Calc. for 2H20, 9.2 per cent.). It melts to a turbid liquid a t 200° (corn.) which does not become clear until 224O a t which temperature effervescence begins. C,H,02N,,HN03 (189.1) requires C = 31.7 ; H = 3.7 per cent. Found in salt dried a t looo N=19*4. 2-Methylglyoxaline p'crute crystallises from boiling water iB fine Found N = 22.3. 2-Methylglyoxdine hydrogen oxnlate crystallises from water in large rhombic prisms which contain 2H,,O (Found H20 = 17.6. Calc. for 2H20 H20=17.3 per cent.). After drying a t looo it melts a t 160° (curr.) and efferv'esces on further heating.Ib is much more readily soluble in water than the corresponding glyoxaline salt. C5H,02N,,C,H30,N (355.2) requires N = 19.7 per cent. needles which are anhydrous and melt a t 2 1 3 O (corr.). C,H6N2,C,H,0,N (311.2) requires N = 22.5 per cent, Found in dried salt W=16.1. C4H,N2,C2H204 (172.1) requires N = 16.3 per cent. 2-Ethylglyoxali~~e-4 5-dicarboxylic Acid. This acid was prepared in the same way as the methyl substituted acid. From 32 C.C. of propaldehyde and the nitrotartaric acid obtained from 50 grams of tartaric acid 43 g r m s of hydrated 2-ethylglyoxaline-4 5 -dicarboxylic acid were obtained that is 64 per cent. of the theoretical yield; Maquenne obtained 30 per cent. 2-Ethylglyoxaline-4 5-dicarboxylic acid melts and effervesces a t 2 5 9 O (cotr.).L 232 FAROEER AND PYMAN: 2-PhenylgEyoxaline-4 5-dicarboxylic Acid. The nitrotartsric acid from 25 grams of tartaric acid was treated with 100 C.C. of aqueous ammonia in the manner previously described. Then 20 grams of benzaldehyde were added with stirring below Oo and the stirring was continued for seven hours, the temperature of the mixture being gradually allowed to approach that of the room. After keeping overnight 17.1 grams of 2-phenylglyoxaline-4 5-dicarboxylic acid were isolated that is, 48 per cent. of the theoretical yield whereas Maquenne’s yield was only 8 per cent. 2-Phenylglyoxaline-4 5-dicarboxylic acid melts and effervesces at 271O (corr.). When distilled under the conditions previously described in the case of glyoxaline-4 5-dicarboxylic acid it gives 2-phenylglyoxaline in a yield of more than 80 per cent.of the theoretical. 2-Phenylglyoxaline crystallises from water in small prismatic needles which melt a t 148-149O (corr.) and are anhydrous. 2-Phenylglyoxaline nitrate is readily soluble in water but less so in alcohol from which it separates in leaflets containing $H,O, which is lost a t 60° in a vacuum. The dried salt melts a t 135O (corr.). Found in air-dried salt H20=6-1; in dried salt N=20*0. C,H8N2,HN0 (207.1) requires N = 20.3 per cent. The hydrogen oxalate crystallises from water in flattened needles which melt and effervesce a t 219O (corr.) and are anhydrous. It is readily soluble in hot water but less so in cold. Found N ~ 1 2 . 0 . CgH8N,,C2H204 (234.1) requires N = 12.0 per cent.The picrate is sparingly soluble even in boiling water from which it crystallises in fine needles which melt a t 238O (corr.) and are anhydrous. Found N = 18.6. CgH8N2,C,H,0;N3 (373.2) requires N = 18.8 per cent, 4 5-Dijmethyl- and 2 4 5-Ti.irneth~yFglyoxaline. When 4 5-dimethylglyoxaline is prepared by Windaus’ method (Ber. 1909 42 758) it is contaminated with 2 4 5-trimethyl-glyoxaline which results from the interaction of diacetyl and ammonia (von Pechmann Ber. 1888,21 1414). 8.6 Grams of diacetyl were dissolved in 50 C.C. of water 50 C.C. of 40 per cent. aqueous formaldehyde added the mixture cooled to Oo and 80 C.C. of concentrated ammonia solution graduall NITRO- ARYLAZO- AND AMINO-OLYOXALINES.233 added the reaction mixture being stirred and kept below Oo. After the addition was ended the mixture was allowed to remain in a cool place overnight then evaporated to a low bulk saturated with anhydrous potassium carbonate and the oil which separated extracted by ether. The crude extract which was contaminated with hexamine amounted t o 5.9 grams. After destruction of the hexamethylenetetramine by boiling dilute hydrochloric acid the picrates of the constituent bases were fractionated from water, when 5-7 grams of 4 5-dimethylglyoxaline picrate (17.5 per cent. of the theoretical yield) were obtained first and then 3.5 grams of 2 4 5-trimethylglyoxaline picrate. 2 4 5-Trimethylglyoxali~e picrate sinters from 160° and melta a t 1 6 3 O (corr.). It crystallises from water in well-defined prisms, which are often serrated.Found N=20*6. C6H,,N2,C6H,0,N requires N = 20-6 per cent. The hydrochloride previously prepared by von Pechmann, crystallises from alcohol in fine needles which are anhydrous and melt at 316O (corr.) (Found N=19.0; C1=24.2. Calc. N=19*1; C1= 24.2 per cent.). 4 5-Dimethtylglyoxaline hydrochloride crystallises from alcohol in well-defined rhombic prisms which melt and decompose a t 305O (corr) . Found N = 21.1 ; Cl= 26.4. C,H,N,,HCl requires N = 21.1 ; c1= 26-7 per cent. 4 5-Dimethylglyoxaline was also prepared by a modification of Kunne's method (Ber. 1895 28 2039; compare also Jowetti T., 1905,87 407). Nine grams of methyl a-isonitrosoethyl ketone were reduced with stannous chloride as described by Kunne but the temperature of the reaction mixture was maintained a t 1 5 O and, after the removal of the tin the evaporation of the liquor was conducted entirely under diminished pressure.By these means, a yield of 10 grams of crude crystalline methyl a-aminoethyl ketone hydrochloride was obtained as against 4.2 grams of syrup obtained by Kunne. When this product was heated on the water-bath for four hours with 10 grams of potassium thiocyanate and 40 C.C. of water 5.2 grams of 2-thiol-4 5-dimethylglyoxaline separated and this gave 4 5-dimethylglyoxaline picrate in a yield of 85 per cent. of the theoretical when oxidised with the calculated quantity of ferric chloride.* The yield of 4 5-dirnethylglyoxaline from methyl ethyl ketone is thus 23.8 per cent.of the theoretical. The method of oxidising thiolglyoxalines to glyoxalines by means of ferric chloride has been deecribed by one of:rlis (T.j 1911 99 2176) in the CMQ o 234 FARGKER AND PYMAN: Part 11. N i t r o g l y o x n l i n e s . Rung and Behrend ( l o c . cit.) prepared 4-nitroglyoxaline in a yield of 36 per cent. of the theoretical by boiling glyoxaline with a mixture of nitric and' sulphuric acids. The yield can be improved greatly by the method given below. Eight grams of glyoxaline were dissolved in 16 C.C. of nitric acid (D 1-4) cooled and 16 C.C. of sulphuric acid cautiously added. A vigorous reaction ensued, and when this had subsided the mixture was boiled gently for two hours allowed to ml and then poured into icewater when 7.85 grams of 4-nitroglyoxaline separated.The mother liquors yielded a further 0-5 gram of 4-nitroglyoxaline identical with the above, but no glyoxaline and merely a trace of other crystalline material. The total yield of 4-nitroglyoxaline thus amounted to 63 per cent. of the theoretical. 4-Nitroglyoxaline cryshllises from boiling acetic acid or from alcohol in stout rhombio prisms which are anhydrous and melt a t 312-313O (corr.) (Found N=36*8. Calc. : N=37.1 per cent.). It is only very sparingly soluble in boiling water. Although it dissolves in strong mineral acids it is pre-cipitated unchanged on the addition of water and is recovered unchanged when crystallised f r m aqueous picric acid. 4-Nitl.cr-2rnethylglyoxccline was similarly prepared. ItL crystal-lises from water in fine needles which are anhydrous and melt a t 254O (corr.) sintering from 251O.Found N=33.0. C,H,0,N8 (127.1) requires N = 33.1 per cent. 5-Nit~o-4-methylglyoxalinle was prepared by Windaus (Zm. c i t . ) in a 60 per cent. yield by warming 4-methylglyoxaline with fuming nitric acid a t 80°. Using this method we found the main pro-duct to be 4-rnethylglyoxaline nitrate. Proceeding according to the method described for 4-nitroi-2-methylg1yoxaline 5 grams of 4-methylglyoxaline gave 7 grams of 5-nitro-4-methylglyoxaline (Found N=32*8. Calc. N=33*1 per cent.) melting a t 248O (corr.) that is 90 per cent. of the theoretical yield. 2-thiol-4-aminomethylglyoxaline. The low yield of 4-aminornet~ylglyoxdine recorded (56 per cent. of the theoretical) was due to the fact that insufficient ferric chloride had been employed.When the calaulated quantity (16.2 grams) of this reqpnt is med the product is obtained in a yield of 90 per cent. of the theoretical NTTRO- ARYLdeO- AND AMINO-GLYOXALINES. 236; Attempted Nitratioru of 4 54?im$hylglyoxdh~e. To five grams of 4 :5-dimethylglyoxaline dissolved in 15 C.C. of nitric acid (D 1*4) 15 C.C. of sulphuric acid were added. The first vigoraus reaction was controlled by cooling and after it had ended the mixture was heated for two hours on the water-bath. From the reaction product 1-7 grams of 4 5-dimethylglyoxaline were recovered together with 0.3 gram of the nitrate af 4-metbyl-glyoxaline-5-carboxylic acid (Found C = 32.1 ; H = 4.0 ; N = 21.7. Calc.C=31.6; H=3.7; N=22-1 per cent.) which deposited the corresponding acid melting and effervescing at 222O on the addi-tion of the calculated quantity of sodium hydroxide. From the pure acid the hydrochloride which melted and decomposed a t 231° and the nitrate which deccunposed a t 1 8 9 O were prepared. The melting points of the acid and its salts are in agreement with those found by Gemgross (ZOG. cit. j for 4-methylglyaxaline-5-carboxylic acid. Reduction of Nitroglyoxalines with Tim and Hydrochloric Acid. When 4-nitroglyoxaline 4-nitro-2-methylglyoxaline or 5-nitro-4-methylglyoxaline is reduced with tin and hydrochloric acid and the product mixed with sodium hydroxide and distilled into standard acid two of the three atoms of nitrogen present in the molecule are eliminated in the form of ammonia: 0.5609 of 4-nitroglyoxaline gave 0.1746 NH,; calc.as above, 0-4292 of 4-nitrrv-2-methylglyoxaline gave 0.118 NH ; cale. as 0.4931 of 5-nitro-4-methylglyoxaline gave 0.1378 NH ; calc. as That the greater part of the ammonia is actually produced during the reduction and not by the subsequent action of the alkali is shown in the case of 4-uitroglyoxaline by the following experiment . Twelve grams of 4-nitroglyoxaline were reduced by means of tin and hydrochloric acid in the usual manner. The reduced liquors were freed from tin and then evaporated to dryness then moistened with alcuhol and again evaporated to remove water as far as possible. The crude product was extracted with alcohol and left 9 grams of a crystalline solid which proved taa be ammonium chloride (Found N = 26.9 ; C1= 66.0.Calc. N = 26.2 ; 0.1688. above 0.115. above 0.1320 236 FARGHER AND PYMAN: Cl=66*3 per cent.). The residue of the purple alcoholic solution gave 5 grams of an insoluble phosphotungstate after the removal of ammonia. This produet has not yet been investigated. Reduction of Nitroglyoxalines with So&um Hyposulphite. Behrend and Schmitz (loc. cit .) observed that 4-nitroglyoxaline gave a beautiful blue dye when treated with alkaline reducing agents. We can confirm this result. but find that ammonia is also produced in an amount corresponding witlh the loss .of two atoms of nitrogen in this form from three molecules of 4-nitroglyoxaline when this compound is reduced with sodium hyposulphite in aqueous sodium hydroxide : 0.5148 of 4-nitroglyoxaline gave 0.0521 NH,; calc.as above, The liquors remaining from the distillation gradually acquired a dark blue colour on exposure to the air and on acidification with acetic acid deposited rather less than 0.1 gram of a blue compound which did not melt below 300O. The reduction of 5-nitro-4-methylglyoxaline with alkaline sodium hyposulphite led to t.he same result as in the case of 4-nitro-glyoxaline two molecules of ammonia being produced from three molecules of the nitro-compound (0.5311 gave 0.0487 NH ; calc. as above 0,0474). The reduced solution gradually acquired a rose colour on exposure to air but gave no precipitate with acetic acid. 4-Nitro-2-methylglyoxaline behaved differently from the above compounds on reduction with alkaline sodium hyposulphite yield-ing one molecule of ammonia from three molecules of the nitro-compound (0.5084 gave 0.0230 NH,; calc.as above 0.0227). 0.0516. Part I I I . Ary l a x o g l y o x a l i ~ a e s . 2-Benzeneazoglyoxaline (IV p. 221). 23.25 Grams of aniline were dissolved in 62.5 C.C. of hydro-chloric acid and 187.5 C.C. of water and diazotised wibh 18 grams of sodium nitrite dissolved in 100 C.C. of water. The solution was run slowly into a well-stirred solution of 17 grams of glyoxaline and 40 grams of anhydrous sodium carbonate in 1250 C.C. of water, previously cooled to 5O and kept overnight. The insoluble orange powder was collected washed well with water and extracted suc-cessively with 250 125 and 125 C.C.of cold 2.5 percent'. hydrochloric acid. (Extract= A .) The insoluble malerial amounted to 4. NITRO- ARYLAZO- AND AMINO-GLYOXALINES. 237 grams and after crystallisation from alcohol gave 2 4 5-tris-benzeneazoglyoxaline of which only 0.5 gram was obtained in a pure state. This compound decomposes a t about 200° and effervesces a t 208O (corr.). When pure it is only sparingly soluble even in boiling alcohol from which it crystallises slowly in dark brown clusters of crystals of indeterminate shape. (0.84 required 60 C.C. of boiling alcohol.) Found C=66.0 ; *H =4.6 ; N=29*0. C,,HlGN8 (380.3) requires C = 66.3 ; H = 4-2 ; N = 29.5 per cent. Trisbenzeneazoglyoxaline is insoluble in cold dilute hydrochloric acid and is decomposed when boiled with this reagent.It dis-solves to some extent in aqueous sodium hydroxide. The hydrochloric acid extract ( A ) was diluted with water and basified with sodium carbonate when crude 2-benzeneazoglyoxaline was obtained. as a yellow crystalline precipitate which after thorough washing with cold water and drying amounted to 34 grams. On crystallisation from 150 C.C. of alcohol 31 grams of the pure base were obtained that is 74 per cent. of the theoretical. No other definite compound could be isolated from the mother liquor. 2-Benzeneazoglyoxd~ne crystalliees from alcohol in large orange tablets resembling potassium dichromate in appearance. It melts a t 190° (corr.) to a reddish-black liquid. Found C=62*7 62.7; H=4*8 4.9; N=32*3 C,H,N (172.1) requires C = 62.8; H = 4.7 ; N = 32.6 per cent.Rzhg and Behrend’s Method.-By this method in which benzene-diazonium chloride is allowed to react with glyoxaline without the addition of alkali 5 grams of glyoxaline gave 3.3 grams of crude precipitate insoluble in water. Of this 2.2 grams were separated into 0.7 gram of insoluble resin which appeared to evolve gas on keeping and 1*45 grams soluble in acid which gave pure 2-benzeneazoglyoxaline on crystallisation from alcohol. The crude precipitate was less readily purified by direct crystallisation from alcohol. General PropeT t ies of A r y laz ogl y oxalimes .-To avoid repetition, it will be convenient to describe the general properties of the monoarylazoglyoxalines a t this point. 2-Benzeneazoglyoxaline and 2-benzeneazo-4-methylglyoxaline are fairly readily soluble in alcohol ethyl acetate or acetone sparingly so in ether chlore form or benzene.5-Benzeneazo-4-methylglyoxaline 2-p-bromo-benzeneazoglyoxaline and 4-~bromobenzeneazo-2-methylglyoxaline are sparingly soluble in the first three solvents and very sparingly so in the last three. L 238 FARGHER AND PYMAN: These compounds are almost insduble in cold water or in dilute aqueous ammonia or sodium carbonate but dissolve to some extent in dilute aqueous sodium hydroxide. The benzeneazo-compounds dissolve readily in dilute hydrochloric acid and the solutions yield crystalline hydrochlorides on concentration ; the hydrochlorides of the p-brormobenzeneazo-compounds are sparingly soluble in water. The stability of a 2- and a 4-substituted member of the group towards boiling dilute hydrochloric acid was examined.When 0.5 gram of 2-benzeneazoglyoxalLne was boiled with 20 C.C. of 10 per cent. aqueous hydrochloric acid for two hours under a reflux condenser 0.35 gram was recovered little changed on the addition of ammonia and readily gave the starting material in a pure state on crystallisation from alcohol. When 5-benzeneazo-4-methylglyoxaline was boiled with an excess of 10 per cent. aqueous hydrochloric acid for a few minute it was recovered unchanged after the addition of ammonia but after boiling for one hour it was mainly decomposed with the form-ation of resinous compounds. The arylazoglyoxalines dissolve in concentrated sulphuric acid, giving bright - col ou red soh ticrns.The mono a ry 1 azed erivativ e ~ s yield mainly orange or magenta solutions the 2-substituted deriv-atives being more intensely coloured than the 4-substituted com-pounds whilst the solutions of bis- and tris-arylazoglyoxalines are green and still more intense than those of the Z-monoarylazo-derivatives. Reduction of 2-Benzeneasoglyoxaline 'With Stannous Chloride : Zgolation of 2-Amino-4-p-aminophenylglyoxdhae 2-Amino-glyoxalke Gumaidhe and Aniline. Twenty grams of 2-benzeneazog1yoxaline were dissolved in 200 C.O. of boiling 2.5 per cent. hydrochloric acid and mixed with 120 C.C. of stannous chloride solution.* The solution was immedi-ately dwulorised and when mixed with ZOO C.C. of hydrochloric acid deposited a crystalline tin salt ( A ) . This was collected and the mother liquor was evaporated to dryness dissolved in hot water, and freed from tin.It was then evaporated to low bulk mixed with sodium carbonate and extracted with ether which removed 3.0 grams of crude aniline. The alkaline liquor was acidified faintly with hydrochloric acid evaporated to dryness and extracted with alcohol when 3.1 grams of extract were obtained. This was * The sltannoua chloride solution employed throughout this investigation waa mad6 by mising 40 grams of " tin salt " with sufficient hydrochloric mid to make 100 c.c.. of solution MTRO- ARYLAZO- AND BMINO-QLYOXAUNES. 239 mixed with stannio chloride and deposited first 2.2 grams of pure 2-aminogtyoxdine stanniehloride then crops of the crude salt f r m which a further quantity of 1.0 gram of the pure*salt was obtained, the total yield amounting t o 11 per cent.of the theoretical. The final stannichloride mother liquors were deprived of tin by means of hydrogen sulphide and mixed with picric acid. After crystal-lisation from water the first crop of picrate which melted a t 325O, was decomposed by sulphuric acid the picric acid being removed by means of ether. The solution of sulphates was deprived of sulphuric acid by barium hydroxide and from excess of this reagent by carbon dioxide. The resulting solution was neutralised with aqueous oxalic acid mixed with as much more aqueous oxalic acid and concentrated when crude guanidine hydrogen oxalate separated. After recrystallisation from water this amounted to 0-07 gram melting and effervescing a t 172-173O (corr.) alone or when mixed with pure guanidine hydrogen oxalate.The crystalline tin salt ( A ) was dissolved in water treated with hydrogen sulphide filtered from tin sulphide and concentrated, when 18.55 grams of 2-amino-4-p-aminophenylglyoxaline d i h y h o . chZoride separated that is 64.6 per oent. of the theoretical yield. 2-Am.ino-4-p-aminophelzylg.lyozal~~e C,H,,N (VI p. 222). When the dihydrochloride is mixed with an equivalent quantity of sodium carbonate a colourless oil separates which solidifies on keeping. This is a carbonate for it effervesces on treatment with acid and when dissolved in boiling .water disengages carbon dioxide vigorously on the addition of animal charcoal leaving a solution of the free base which crystallises on keeping.This solution becomes brown a t the top owing to oxidation in the air whilst the larninrr become mauve where exposed to the light. To 5-0 grams of the dihydrochloride in 50 C.C. of boiling water, 30 C.C. of hot 10 per cent. aqueous sodium carbonate and a pinch of animal charcoal were added. The solution was boiled for five minutes filtered and kept when 3.1 grams of the base separated and were recrystallised from water. This base cryshllises from water in glistening leaflets which melt and effervesce at 148O (corr.). It contains 1H,O which is not l.ostl in a vacuum or on heating a t looo. Found C= 56.3 ; H = 6.4 ; N = 29.5 29-2. C,H,,N4,H,0 (192.2) requires c f = 56.2 ; H = 6.3 ; N = 29.2 per cent. It is sparingly soluble in cold fairly readily so in hot water; fairly readily soluble in cold readily in hot alcohol and very sparingly so in chloroform or ether.L' 240 FARCHER -4ND PYMAN: An aqueous solution of the base gives with silver nitrate a white precipitate which blackens at once on the addition of ammonia; with Fehling's solution a nearly black precipitate-presumably a copper salt-which is unchanged on boiling the solution ; with cold permanganate instant reduction ; with sodium diazobenzene-p-sulphonate in aqueous sodium carbonate an immediate cherry-red colour. When the base is dissolved in an excess of hydrochloric acid and mixed with sodium nitrite a yellow solution is obtained, which yields with a solution of &naphthol in aqueous sodium hydr-oxide a sparingly soluble purple dye.On the addition of sodium hydroxide to a solution containing 2-amino-4-p-aminophenyl-glyoxaline hydrochloride and sodium nitroprusside a green color-ation changing to1 chestnut-brown is produced. On the addition of dilute sulphuric acid to an aqueous solution of the base or its hydrochloride the very sparingly soluble sulphate crystallises in woolly needles. The dihydrochloride crystallises from dilute hydrochloric acid in colourless prisms which do not melt below 3QOO. It is readily soluble in cold very readily so in hotl water, Found C1= 28.6 ; N = 22.5. The dipkrate forins yellow silky needles which darken a t 245' It is very sparingly soluble even The benzylidene derivative of 2-amino-4-11-aminophenylglyoxaline C,Hl,N4,2HCl (247.1) requires C1= 28.7 ; N= 22.7 per cent.and decompose a t 250° (corr.). in boiling water. was not obtained in a crystalline form. 2- A ce tylamino-4-p-ace tylnmin o pkenylgly oxdin e . 10.6 Grams of 2-amino4-p-a1ninophenylglyoxaline were boiled with 50 C.C. of acetic anhydride for one hour under a reflux con-denser and mixed with aqueous sodium carbonate when 13.9 grams of the diacetyl derivative were obbained that- is 98 per cent. of the theoretical yield. The base forms a colourless crystalline powder which does not melt below 300O. Found N = 21.2. It dissolves in dilute hydrochloric acid but the hydrochloride crystallises almost a t once. It appears to be changed by prolonged boiling with hydrochloric acid, The hydrochloride was consequently prepared by triturating the base with an excess of 10 per cent.aqueous hydrochloric acid drain-ing the insoluble salt? and crystallising it from water when it C,,H,,O,N (258.2) requires N = 21.7 per cent NITRO- ARYTJAZO- AND AMINO-Q~~YOSALTNES. 241 formed colourless prismatic needles which did not melt below 300°. It is sparingly soluble in cold fairly readily so in hot water. Found in air-dried salt loss a t l l O o = l l * l . Found in salt dried a t l l O o C=53.1; H=5-3; N=18*7; C,,H,,O,N,,HCl (294.7) requires C=53*0; H=5*1; N=19.0; (31 = 12.0 per cent. Oxidation .-Ten grams of 2-acet.ylamina-4-p-acetylaminophenyl-glyoxaline were suspended in 150 C.C. of cold water and mixed with 4 grams of 50 per cent. aqueous sulphuric acid when a suspension of the sulphate resulted.To this cold 4 per cent. aqueous potassium permanganate was added until a test portion of the pfo-duct remained pink for a few seconds about 240 C.C. being required. The liquor was then filtered from manganese hydroxide acidified with hydrochloric acid and extracted with ether. The et-hereal extract amounted to 1.5 grams and after digestion with a litt81e warm water left 1.0 grain of p-acetylaminolbenzoic acid which melted at 260° (corr.). After recrystallisation from boiling water, t-he acid formed glistening needles having the same melting point. A specimen of the pure acid from another source and a mixture of the two melted a t the same temperature. The identification was confirmed by analysis (Found C - 59.9 ; H = 5.2 ; N = 7-8. Calc. : C=60*3; H ~ 5 .1 ; N=7*8 per cent.) and by hydrolysis to p-amino-benzoic acid which melted a t 190° (corr.) alone or mixed with t h e acid resulting from the reduction of pnitrobenzoic acid. ClsH,,0,N,,HC1,2H20 (330.7) requires 2H,O = 10.9 per cent. Cl= 11.6. Redzcctiom of 2-Benzeneaxogl~oxal~ne with Zinc Dust and Acetic Acid Isolation of Glycocyarnidine Aniline a d 2-Arnitto-4-p-amino phemylglyoxdine . To a boiling solution of 17.2 grams of 2-beazeneazoglyoxaline in 100 C.C. of glacial acetic acid and 300 C.C. of water 45 grams of zinc dust were added gradually in the course of twenty minutes without further heating. The excess of zinc was removed the liquor diluted with 2 litres of water giving an indigecoloured solution and treated with hydrogen sulphide. After collecting the zinc sulphide-which had carried down t.he colouring matter-the liquor was mixed with 20 C.C.of hydrochloric acid and evaporated to dryness. The residue was dissolved in a little water mixed with sodium carbonate and extracted with ether when 2.7 grams of insoluble black material were deposited ; this contained zin 212 PAWKER AND PYMAN: carbonate and the carbonate of 2-amino-4-p-aminophenylglyoxaline. The ethereal extract left on evaporation 7.0 grams of practically pure aniline. The alkaline liquor was acidified faintly with hydro-chloric acid mixed with a solution of 23 grams of picric acid in 1 litre of boiling water and stirred when 2.6 grams of 2-amino-4-paminophenylglyoxaline dipicrate separated immediately as a brownish-yellow crystalline powder which melted a t 240"; for the identification of this substance the hydrochloride and base were prepared and found to have the properties recorded above.The filtrate from this salt was kept overnight when 21.5 grams of a granular crystalline picrate melting at 196O separated and on concentrating the mother liquor a further 4.8 grams melting a t B O O were obtained. These crops were mixed converted into the hydrochloride and crystallised from alcohol whm eventually 5.9 grams of pure glycocyamidine hydrochloride were obtained that is, 43 per cent. of the theoretical yield. It formed clusters of pris-matic needles which began to darken and sinter a t 205O and melted a t 211-213O (corr.). E. Schmidt (Arch. Phmm. 1913, 251 557) states that it begins to discolour at 200° and melts a t 208-210° (Found C= 26.5 ; H = 4.4 ; N = 30.6 ; C1= 26.3.C3H,0N3,HC1 (135.6) requires C = 2 6 * 6 ; H=4.5; N=31-0; C1= 26-2 per cent.). To complete the identification of this compound the base and some other salts were prepared. The base crystallised from water in colourless prismatic needles which began to darkea slowly from about 220° and quickly from about 250° without melting even a t 300O. It was anhydrous. (Found C=36*4; H=4*8; N-42.3. C3H,0N (99.1) requires C= 36.3 ; H -5.1 ; N = 42.4 per cent.) E. Schmidt (Zoc. c i t . ) states that glycocyamidine darkens from 220°, but does not melt a t 250O. It gave with sodium nitroprusside and sodium hydroxide an orange solution which became Burgundy-red on the addition of acetiu acid (Weyl's reaction).It is stable towards cold aqueous permanganate in acid solution but reduces cold alkaline permanganate yielding a green solution. The platinichloride was obt'ained on spontaneous evaporation of an aqueous solution in large transparent quadrilateral tablets having the composition C3H,0N3,H2PtC1,,2H,0. It begins to darken a t 220° gradually sinters and is quite black by 260* with-out actually melting even a t 300O. E. Schmidt (loc. &.) found that glycocyamidine platinichloride had this composition and did not melt a t 260° but sintered and blackened earlier. The platinichloride was also obtained in an anhydrous form by crystallisation from a hot concentrated solution when it formed clusltere of prisms NITRO- ARYLAZO- AND AMINO-QLYOXALINES.243 The additive compound with gold chloride C3H,0N8,AuC&, melted a t 157-158O (corr.). Eorndorfer (Arch. Phum. 1904, 242 633) found that glycocyamidine gold chloride had this com-position and melted a t 153-154O. The picrate crystallised from water in glistening striated yellow leaflets (flat needles) which melted a t 215-216O (corr,). Jaff6 (Zeitsch. physiol. Chem. 1906 48 430) describes glycocyamidine picrate as forming needles which melt at 210O. 2- and 4-p-Bromobenzeneaaoglyoxaline. 34.4 Grams of p-bromoaniline i n 200 C.C. of hydrochloric acid and 600 C.C. of water were diazotised a t -2O to Oo by a solution of 14-4 grams of sodium nitrite in 72 C.C. of water. The solution was kept for twenty minutes and poured in a slow stream into a solu-tion of 13.6 grams of glyoxaline and 300 grams of sodium carbonate crystals in 2 litres of water previously cooled to 5O.After adding a little more aqueous sodium carbonate the mixture was kept over-night and the insoluble yellow powder collected and washed with water. It amounted to 48.7 grams after drying in the air decom-posed a t 245O and was almost completely soluble in dilute hydro-chloric acid After fractional crystallisation from alcohol there were obtained 37.5 grams of pure 2-p-bromobenzeneaaoglyoxalil.L~ and 5.1 grams scarcely less pure whilst the final mother liquors deposited a mixture of this compound with dark brown w a h , which were separated mechanically and amounted to about 2.5 grams melting atL about 175O. These were dissolved in dilute hydrochloric acid and the solution was filtered from a little dark brown insoluble matter and mixed with ammonia when a yellow, gelatiinous precipitate was formed which readily became crystalline on warming and stirring.This base was collected and crystallised several times from alcohol when 4-p-bromo benzeneazoglyoxali~e was obtained in a pure state. 2-p-Bromob enaeneazoglyoxaline crystallises from alcohol in chestnubbrown prismatic needles which melt and decompose at) 253O (WIT.). Found C = 42.9 ; H = 3.1 ; N = 22.1. C,H,N,Br (251.1) requires C =43*0 ; H = 2.8; N = 22.3 per cent. 4 -p -Brmob enz eneaz ogly o xalin e cry st allises from alcohol in-clusters of brownish-yellow prisms which melt and decompose a t 191O (corr.). Found C=43*2; H=3*1; N=21*9.C,H,N4Br (251.1) requires C=43*0; H=2.8; N=22’3 per cent 244 FARCTHER ANT) PYMAN: On reducing this base (0.26 gram) with stannous chloride and distilling the resulting solution with an excess of sodium hydroxide, the volatile products consisted of pbromoaniline (0.18 gram) and ammonia which gave 0.05 gram OE ammonium chloride. Reduction of 2-p -Br o mo b en z en eaz og ly oxdine with S t ann ow s Chloride Isolation of 2-Aminoylyoxaline p-Bromoaniline, Gwnidine 2- A mino - 4 - p-amin o phe n y lgl y o xalin e and a Base, C,H,N,Br-To 78 grams of 2-p-bromobenzeneazoglyoxaline suspended in 40 C.C. of hydrochloric acid and 1 litre of boiling water 400 C.C. of stannous chloride solution were added. The solution immedi-at8ely became decolorised and after the removal of 0.5 gram of brown insoluble matter was concentrated under diminished pressure.The tin salts which separated were coWected from time t a time and combined so that the product was obtained in two fractions consisting of the crystalline tJn salts and the syrupy residue. The crystalline tin salts were dissolved in water and deprived of tin by hydrogen sulphide. The solution of hydrochlorides was evaporated to dryness dissolved in a little water and mixed with aqueous sodium carbonate when 39.6 grams of p-bromoaniline separated. The filtrate from this gave a further 1.1 grams of the same compound on extraction with ether and was next acidified with hydrochloric acid evaporated to dryness and extracted with absolute alcohol. (Insoluble material = A .) On distilling the alcohol a brown syrup remained which quickly crystlallised and became a rock-like mass of 2-aminoglyoxaline hydrochloride, amounting t o 15.8 grams and melting a t 135-140°.The syrupy tin salts were also dissolved in water and deprived of tin. The resulting solution was evaporated to dryness dissolved in a little water mixed with sodium carbonate and extracted with ether. This on concentration deposited 0.9 gram of colourless needles melting at 1 7 8 O (corr.) which proved to b e 8 base having the composition C9H,N,Br (compare p. 245). The ethereal mother liquor on evaporation left 2.6 grams of dark brown syrup which gradually crystallised and consisted largely of p-bromoaniline. The alkaline liquor was acidified with hydrochloric acid, evaporated t o dryness under diminished pressure and extracted with absolute alcohol.The insoluble salts consisting mainly of sodium chloride were combined with those obtained previously ( A ) dissolved in water and mixed with aqueous picric acid when 2.7 grams of 2-amin~4-p-aminophenylglyoxaline dipicrate melt NITRO- ARYLAZO- AND AMINO-QLYOXAT~TNES. 246 ing a t 240° separated; the identity of this salt was confirmed by its conversion into the hydrochloride and base. The alcoholic extract was evaporated under diminished pressure and left 13.0 grams of brown syrup which crystallised only partly on seeding with 2-aminoglyoxaline hydrochloride. It was converted into the stannichloride and crystallised fractionally from 10 per cent. hydro-chloric acid when 10.2 grams of 2-aminoglyoxaline stanni-chloride melting a t 280° (corr.) were obtained.This is equivalent to 4.9 grams of 2-aminoglyoxaline hydrochloride the total yield of which was therefore 20.7 grams that is 56 per cent'. of the theoretical. The remaining stannichlorides were not readily purified by frac-tional crystallisation and were reconverted into hydrochlorides, which amounted to about 3 grams. This material waq mixed with sodium carbonate evaporated until nearly dry and extracted with hot alcohol.* The extract was distilled and the residue mixed with ail excess of 10 per cent. aqueous oxalic acid when 1.6 grams of guanidine hydrogen oxalate separated in large crystals. After recrystallisation from water this salt formed colourless spears, which melted a t 173-174O (corr.) after drying a t looo and was sparingly soluble in water.It had the campmition CH5N3,C2H204,H,0 previously recorded by Strecker (Amden 1861 118 160). (Found H,O=10.3. Calc. H,O=10*8. Found in dried salt C=24-0; H z 4 . 9 ; N=28*0. Calc. C=24.1; H=4*7; N=28*2 per cent.) The melting point of a specimen of guanidine hydrogen oxalate pre-pared synthetically and that of a mixture of the two preparations was the same. The identification was confirme'd by the prepara-tion of the nitrate and picrate which had the properties previously recorded. The base melt'ing a t 178O (corr.) obtained as a by-product in the above reaction (compare p. 244) forms colourless needles from alcohol or ether. It is sparingly soluble in water readily so in cold and very easily soluble in hot alcohol, but sparingly so in ether.Its alcoholic solution gradually becoma purple when exposed to the air. It contains halogen. Found C=43*0; H=3*9; N=21*9. CgHgN,Br (253.1) requires C=42*7; H=3*6; N=22.1 per cent. 0.122 Gram mixed with an excess of hydrochloric acid and evaporated to dryness gave 0.153 gram of salt which is therefore * The method employed for the extraction of guanidine carbonate is unsuitable and it is probable that a considerable proportion remained behind with the sodium carbonate a dihydrachlode (calc. yield 0.167 gram). This salt qstallised from water in elongated leaflets which after drying a t 1000 melted and decomposed a t 245O (oorr.) after sintering earlier. The (&)pimate crystallises from water in woolly needlee which melt a t 225O (corr.) and are sparingly soluble in hot very sparingly so in cold water.The base decolorises potassium permanganate instantly in cold dilute sulphuric acid solution and gives a Burgundy-red coloration with sodium diazobenzene-psulphonate in aqueous sodium carbonate. When dissolved in dilute hydrochlorio auid and mixed with sodium nitrite it yields a colourless crystalline precipitate, but the product-crystals and mother liquor-when poured into alkaline &naphthol gives no coloration. When an aqueous solution of the hydrochloride is mixed with sodium acetate and benzaldehyde a turbid yellow solution is pro-duced-evidently awing to the f ormation of a benzylidene compound. The compwition and mode of formation of the baseindicate that it is 2-p-bromobenzenehydrazoglyoxaline o r a substance resulting from this by the benddine or semidine change.The formation of a dihydrochloride and a benzylidene derivative rule out the first suggestion whilst the formation of the latter compound also eliminates the semidine-type formula I11 given below. This formula and the benzidine-type formula I are also incompatible with the behaviour of the compound on treatment with nitrous aaid and sodium 8-naphthoxide but the semidinetype formula 11, representing 2 -5 I- b romo- Zf-am~n~rtilillag.ly~x~ine admits the possibility of o-diazoimine formation with nitrous acid and is in harmony with all the observed properties of the compound (com-pare p. 223). \-/ (111.) 2-Aminoglyoxali7ts (XII p.223). For the purification of 2-aminoglyoxaline crystallisation of the The free base can be obtained (1) from the hydrochloride by the stannichloide and hydrogen oxalate has proved to be USQfUl NITRO- ARYLAZO- AXD A.WINO-QLYOXALINES. 247 addition of an equivalent quanGty of sodium carbonah evapors-tion to dryness and extraction with alcohol and (2) from the hydrogen oxalate by treatment with aqueous barium hydroxide, removal of the excess of this by carbon dioxide and evaporatio~l of . t h e solution under diminished pressure. I n either case it is obtained as a nearly colourless syrup which gradually turns brown on keeping. It is miscible with water and alcohol sparingly soluble in chloroform but hardly soluble in ether or benzene. The hydrochloride crystallises from absolute alcohol in long, colourless plates which melt a t 162O (corr.).It is deliquescent, and readily soluble in cold very readily so in hot absolute alcohol. Its aqueous solution reacts neutral to litmus. C3H5N3,HC1 (119.5) requires C=30-1; H ~ 5 - 1 ; N=35*1; C1=29*7 per cent. The stanmichloride crystlallises from two la two and a-half times its weight of 10 per cent. hydrochloric acid in prismatic needles, which are anhydrous and melt at 286O (corr.). It is readily soluble in water. Found C=30.2; H=5-2; N=34-7; C1=299. Found @I = 42.4. (C3H5N3),,H,SnCls (499.6) requires C1= 42.6 per cent. The nitrate separates from water in large transparent tablete, which are anhydrous and after drying a t looo sinter from about 125O and melt a t 135-136O (corr.).Found N = 38.2. C,H,N,,HNO (146.1) requires N =38-4 per cent. The hyd!roge.pz. oxalate crystallises from water in large colourless tablets which are anhydrous and melt and effervesce at 2 1 1 O (wrr.). It is sparingly suluble in cold readily so in *hot water. Found N = 24.0. C3H5N3,C,H,04 (1 73.1) requires N = 24.3 per cent. The @crate separates from water in long glistening silky needles or in shortl prismatic needles both melting a t 236O (corr.) after drying a t looo. It is'sparingly soluble in cold fairly readily so in hot water. Reaction8 of 2-&4 m~~~g~yoxal~~ae.-%-Amino~l~oxalin~ hydro-chloride dissolved in dilute aqueous copper sulphate gives oh the addition of sodium hydroxide a green precipitate which rapidly darkens and becomes purple-brown.The same precipitate-evidently a copper salt-is obtained eventually with Fehling's solu-tion; no reduction of this solution takes plaoei even on boiling. 2-Aminoglyoxaline nitrate in aqueous silver nitrate gives a whit 248 FAROHER AND PYMAN: precipitate on the addition of ammonia; this precipitate is soluble in excess of ammonia and the solution deposits metallic silver on heating. 2-Aminoglyoxaline hydrochloride in aqueous solution decolorises aqueous potassium permanganate instantly ; with ferric chloride it gives no coloration. With sodium diazobenzene-p-sulphonate in aqueous sodium carbonate it gives a deep red colour. On the addition of sodium nitrite to aqueous 2-aminoglyoxaline hydro-chloride a clear yellow solution is produced which gives a soluble, brownish cherry-coloured dye with 8-naphthol in aqueous sodium hydroxide.An aqueous solution of 2-aminoglyoxaline hydro-chloride mixed with dilute aqueous sodium nitroprusside gives on the addition of sodium hydroxide a deep blue colour which slowly changes to a bright chestnut on keeping. 2-Aminoglyoxaline is very stable towards hot acids and alkalis. When boiled with 10 per cent aqueous sodium hydroxide no ammonia is evolved and it can be recovered unchanged from the solution. It can be recovered mainly unchanged after heating with concentrated hydrochloric acid for three hours a t 170° and even after three hours at 200° a small proportion can be recovered, together with ammonium chloride and other unidentified products. An aqueous solution of 2-aminoglyoxaline hydrochloride cont.ain-ing an excesj of sodium acetate gives no coloration or other evidence of the formation of a benzylidene derivative when mixed with benzaldehyde .2-8 cetyZnminogZyoxaZine was prepared by bqiling 2-amino-glyoxaline hydrochloride with anhydrous sodium acetate and acetic anhydride for one hour and mixing the product with aqueous sodium carbonate. It crystallises from water in small prisms which melt to a brown liquid at 287' (corr.) after sintering and darken-ing from about 270O.' It is anhydrous and sparingly soluble in cold wat'er but fairly readily so in hot water. Found C= 47.7 ; H = 5.7 ; N = 33-4. C,H,ON (125.1) requires C=48*0; H=5*6; N=33.6 per cent. The reactions of this substance are described with those of the next compound.2 - B e n z o y l a m i i z ~ g l y o x ~ ~ ~ e was prepared by the Schotten-Bau-mann method. The crude product collected from the reaction liquor appears to be a di- or tri-benzoylaminoglyoxaline. After washing with ether t o remove benzoic anhydride it formed a nearly colourless crystalline powder which contained only a trace of chloride but gave an odour of benzoyl chloride when boiled with dilute hydrochloric acid. When treated with a little hot alcohol NITRO- ARYLAZO- AND AMINO-GLYOXALINES. 249 it dissolved and 2-benzoylaminoglyoxal~ne crystallised from the hot liquor whilst the mother liquor from this left an Ql-appar-ently ethyl benzoate-on distillation. 2-Benzoylaminoglyoxaline was purified by crystallisation from alcohol from which it separates in glistening leaflets melting a t 2 2 7 O (corr.) after sintering earlier.It is sparingly soluble even in hot alcohol and almost insoluble in boiling water. Found C = 63.9 ; H = 4.9 ; N = 22.4. C,,H,0N3 (187.1) requires C=64.1; H=4*9; N=22.5 per cent. 2-Acetylaminoglyoxaline and 2-benzoylaminoglyoxaline are soluble in dilute hydrochloric acid and in aqueous sodium hydr-oxide but not in aqueous sodium Carbonate. They give cherry-red solutions with sodium diazobenzene-p-sulphonateA in sodium carbonate but do not give colorations with sodium nitroprusside and sodium hydroxide. They do not change the colour of cold aqueous acid permanganate but give green solutions with cold aqueous permanganate in sodium hydroxide solution.When mixed with hydrochloric acid and sodium nitrite they do not couple with /3-naphthol in aqueous sodium hydroxide. The S e n z e ? i e n z o - 4 - m e t l ~ ~ l ~ l ~ ~ ~ ~ ~ l ~ ? i e s . 37.2 Grams of aniline in 100 C.C. of hydrochloric acid and 300 C.C. of water were diazotised with 28-8 grams of sodium nitrite in 150 C.C. of water. The solution was run slowly into a solution of 32.8 grams of 4-methylglyoxaline and 100 grams of sodium hydrogen carbonate in 2 litres of water a t loo and kept overnight. The orange precipitate was collected washed well with water (filtrate F ) and triturated successively with 500 250 and 250 C.C. of 2.5 per cent. aqueous hydrochloric acid. The insoluble fraction formed a dark red powder which amounted to 23.2 grams and decomposed at 175O after sintering from 1 6 0 O .On crystallisation from 300 C.C. of alcohol it gave 17.3 grams of pure 2 5-bisbenzene-azo-$-met hylglyoxaline the remainder of the material forming a black resin. The hydrochloric acid extract was basified with sodium carbonate, and gave 40.4 grams of a yellow crystalline powder which sintered from 160° and decomposed a t 195O. On crystallisation from 400 C.C. of alcohol it gave successively 13.1 grams melting a t 235O 3.9 grams melting a t 232O which both gave 5-benzeneazo-4-~nethyZglZyoxaZine on recrystallisation then 7.4 grams melting a t 175O which gave 2-benzeneazo-4-methylglyoxaline on recrystallisa-tion then 12.4 grams of a mixture of the two compounds 250 FARQHER AND PYMAX: Owing to the formation of the bis-mmpound in the above r w tion the benzenediazonium chloride employed wits insufficient to combine with the whole of the rnethylglyoxaline present and it was calculated that 10.5 grams of this remained in the filtrate F .This was accordingly treated with a diazo-solution prepared from 11.9 grams of aniline and gave further quantities of tlhe substances described above 5.3 grams of the bis-compound and 5.6 grams of 5-benzeneazo-4-methylglyoxaline being obtained i n a nearly pure S t a b . 2-Bemzeneazo-4-methylglyoxdcline (XIV p. 224) crystallises from alcohol in orange prisms which melt a t 185O (corr.). Found C=64*8; H=5*6; N=30.1. CIOHIONl (185.2) requires C = 64.5 ; H = 5.4 ; N = 30.1 per cent. 5-Benzeneazo-$-met hytglyoxaline (XVIII p. 2 24) crystallises from alcohol in flat glistening copper-coloured needles which melt and decompose a t 240° (corr.).Found C=64*5; H=5-6; N=30*0. C,,H,,N (185.2) requires C = 64.5 ; H = 5.4 ; N = 30.1 p0r cent'. 2 5-Bisb emzeneazo-4-methylgZyoxaline separates from alcohol in Both forms are prismatic needles and from ethyl acetate in cubes. garnet-red in colour and melt and decompose a t 2 0 6 O (corr.). Found C=66.0 65.9; H=5*1 5.1; N=28.8 28.8. ClGHl4NB (290.2) requires C=66.2; H=4*9; N=29*0 per cent. This substance is readily soluble in alcohol ethyl acetate or acetone fairly readily so in chloroform but sparingly so in ether or bpzese. It is soluble in aqueous sodium hydroxide and is reprecipitated unchanged on the additmian of acetic acid. It is only very sparingly soluble in dilute hydrochloric acid.When boiled wit'h 10 per cent. aqueous hydrochloric acid it is quickly resinified with effervescence, doubtless due to nitrogen and the production of an odour of phenol. Reduction of 2-Benzeneazo-4-met hylylyozaline with Stannow ChluritZe . 1.5 Grams of 2-benzeneazo-4-methylglyoxaline gave 1.4 grams of 2 -amino-5-pamin op hen yl-4 -me th ylgl yoxaliiie d ihy drochloride when reduced with stannous chloride i n the manner previously described for the lower hom'ologue (p. 238). dih ydro c hlo&i?e cry&allise~ from water in diamond-shaped plates which are anhydrous and do not melt below 300O. It is readily soluble in cold very readily so in hot water. 2- A n~ i ~ o - 5 - p-amino phenyl - 4 - IIL e t h y lgl y oxali t I NITRO- ARYLAZO- AND AMINO-GLYOXALINES.261 Found C=46-0 45.9; H=5*5 5.5. Cl,H,,N,,2HC1 (201-0) requires C= 46.0 ; H =5*4 per cent. When boiled with an excess of aqueous sodium carbonate and animal charcoal it yields the mono hycEroc?do&#e unlike the lower homologue which yields the corresponding base under this treat-ment. The mumohydrochZoride crystallises from alcohol in flat needles which sinter a t about 80° become discoloured rapidly about 240°, and melt a t 260° (corr.). It is readily soluble in hot water or alcohol less so in these solvente when cold. Found in air-dried base loss a t 60° in a vacuum 13.2 13.3. C,,H&N,,HCl,2~H20 requires loss of 2H20 = 13.4 per cent. Found in base so dried C=51*5; H=5.6; N=24.0 24.0; C,,H’,,N,,HCl,~H,O (233.7) requires C = 51.4 ; H = 6-0 ; N = 24.0 ; C1=15.2 per cent.The &@crate forms glistening yellow needles which melt and decompose a t 255O (corr.) after darkening earlier. It is very sparingly soluble even in boiling water. An aqueous solution of the hydrochloride reduces cold amrnoniacal silver nitrate. It gives with Fehling’s solution a greyish-green precipitate which becomes pale brown on boiling the liquor ; with cold aqueous acid permanganate instant reduction ; with sodium diazobenzene-p-sulphonate a pale mange colour which deepens un keeping; with hydrochloric acid and sodium nitrite an orange-yellow solution which yields a sparingly soluble claret dye when added to a solution of &naphthol in aqueous sodium hydr-oxide. On the additmion of sodium hydroxide t o an aqueous solu-tion of the hydrochloride and sodium nitroprusside an orange colour is produced which changes to green on the addition of acetio acid.The diacetyZ derivative was prepared by the action of sodium acetate and acetic anhydride on the dihydrocbloride and was purified by crystallisation of the hydrochloride. hydro-chloride crystallises from water in felted silky needles which are sparingly soluble in cold water contain 4H,O and after drying a t looo melt and decompose a t 303O (corr.). C1= 14.9. 2-A ce t ylamino-5 -p-acet ylaminoyh enyl-4-me t hylgZy oxuline Foluiid in air-dried salt loss a t looo= 19.0. Found in salt dried at’ looo C1= 11-4. C,,H,,0,N,,HC1,4H20 requires H,O = 18.9 per cent Cl4Hl6O2N4,RCI (308.7) requires C1= 11.5 per cen 2 52.E'ARBHBR AND PYMAN: On adding ammonia to an aqueous solution of the hydrochloride, the base was precipitated in minute glistening needles which after drying a t looo melted to a red liquid a t 280° (corr.). Nono b enzylidefie Derivative .-To 0.5 gram of the dihydro-chloride in 5 C.C. of water there were added first 0.55 gram of sodium acetate in 5 C.C. of water and then 0.5 C.C. of benzaldehyde, and the mixture was stirred. A yellow colour was developed and the aqueous liquor became turbid and gradually deposited crystals. On adding a few drops pf acetic acid and ether the quantity of crystals was increased. They were collected and washed with water and ether when there remained 0.5 gram of a pale yellow, crystalline powder which proved to be the acetate of 2-amino-5-p-b enz ylideneaminophenyl-4-me t hylglyoxnline .When dried a t 1 OOo, it melts and decomposes a t 208O (corr.) after sintering and darken-ing earlier. Found in substance dried in a vacuum C=67*2; H=6*2; N= 16'2." C,,H,,N,,C,H,O (336.3) requires @= 67.8 ; H = 6.0 ; N = 16.7 per cent. This salt is very sparingly soluble in cold water but slightly so in boiling water with which however it gives an odour of benz-aldehyde and thus appears t o suffer hydrolysis. When mixed with aqueous sodium carbonate it yields the base as a deep yellow in-soluble gum which could not be obtained in crystalline form. When the acetate is moistenel with 10 per cent. aqueous hydro-chloric acid it turns red but does not dissolve until the mixture is warmed when the red colour disappears.Reduction of 2-Benzeneazo-4-metkylglyosaline with Zinc Dust and Acetic Acid. Two grams of the azo-compound were reduced by the method applied to the lower hom?logue (p. 241) and worked up in the same manner as far as the removal of the aniline by extraction with ether. The solvent removed 0.65 gram of crude aniline. The alkaline liquor remaining was acidified with hydrochloric acid, evaporated to dryness and extlracted with absolute alcohol when 1.4 grams of brown syrup were removed. This when dissolved in a little absolute alcohol and kept deposited 0.7 gram of nearly pure alacreatinine hydrochloride. This was converted into the picrate when a very small quantity of 2-amino-5-p-aminophenyl-4-methylglyoxaline dipicrate separated * The substance left a trace of m h on combustion NITRO- ARYLAZO- AND AMINO-OLYOXALINES.253 from the hot solution whilst on cooling alacreatinine picrate crystallised out. After recrystallisation the salt was obtained in a pure state and was converted into the base and hydrochloride by the usual methods. Alacreatinine crystallises from water in stout elongated prisms which resemble carbardde and contain lH,O as previously stated by Baumann (Amalert 1873 167 83). After drying a t looo it melts a t 222-223O (corr.). (Found in air-dried salt H,O=13.6. Calc. 13.7. Found in dried salt C=42*4; H=6-3; N=36.9. Calc. cC=42*5 ; H= 6.2 ; N = 37.1 per cent.) It does not give Weyl's reaction and does -not reduce cold aqueous acid permanganate but gives a green solution with cold alkaline permanganate.The hydrochloride crystallises from absolute alcohol in clusters of prisms which are anhydrous and melt a t 202-203O' (corr.). It is very readily soluble in water sparingly soluble in cold fairly readily so in hot' alcohol. Found C1= 23.6. C,H,ON,,RCl (149.6) requires C1= 23.7 per cent. The picrate separates from water in yellow prismatic needles, which are anhydrous and melt and decompose a t 212O (corr.) after sintering from about 200O. It is sparingly soluble in cold fairly readily so in hot water. Found N=24-5. C4H70N3,C,H30,N (342.2) requires N = 24.6 per cent. R edu c t ion of 5 -Ben? en ea z 134-rn e thy l g l y omxdi n e with Stan no us Chloride. Fourteen grams of the azo-compound were dissolved in a boiling mixture of 70 C.C.of 10 per cent. aqueous hydrochloric acid and 140 C.C. of water and mixed with 80 C.C. of stannous chloride solution. The crystalline and residual tin salts were separated as in the experiments described earlier and decomposed separately by hydrogen sulphide. The crystalline salts gave a solution of hydrochlorides which when evaporated nearly to dryness and mixed with alcohol left 6.1 grams of ammonium chloride undissolved. (Alcoholic mother liquor = A .) The residual salts gave a solution of hydrochlorides which on concentration deposited 1.7 + 0.5 grams of the hydrochZm*de C9H,,0N2,HCl described below and on further concentration and addition of alcohol gave 1.5 grams of ammonium chloride. The alcohoIic mother liquor was combine with A and gave 4.7 grams of aniline together with 3-8 gram of a brown gummy hydrochloride.This was a mixture from which only very m a l l quantities of crystalline campaunds were isolated by various methods of treatment. The hydrochloride C,H,,ON,,HCl cryshallism from water in colourless transparent rectangular tablets which melt and effervesce a t 308O (corr.) after sintering and darkening earlisr. It is readily solublel in hot less so in cold water giving a solu6hn which is strongly acid to litmus. Found in air-dried salt# lms a t l10°=l*7. Found in salt dried a t l l O o c'=54.8 54.8 55.0; H=5.9 5.0, C,H,,0N2,H@1 (198.6) requires C = 54.4 ; H = 5.6 ; N = 14.1 ; The correspanding base is obtained by adding ammonia t o a con-centrated aqueous solution of the hydrochloride.It crystallises from water in InIilliant elongated prisms which are anhydrous and melt a t 185O (corr.). Found C=66*6 66.1; €1=6*2 6.1; N=17-8 17.2. C,H,,ON (162.1) requires C= 66.6 ; H = 6.2 ; N = 17.3 per cent. The base is more readily soluble in dilute aqueous sodium hydr-oxide than in water. With silver nitrate it yields a white pre-cipitate which dissolves on the addition of ammonia; on boiling this solution no reduction takes place The base does not reduce Fehling's solution on boiling. It is stable towards cold aqueous acid potassium permanganate, but slowly reduces cold alkaline permanganate giving a green solution. It gives no coloration with sodium diazobenzeae-r)-sulphonate in aqueous sodium carbonate.When dissolved in hydrochloric acid and mixed with sodium nitrite it fails to couple with /3-naphthol in 8queous sodium hydroxide. The hydrochloride is recovered slightly charred but otherwise unchanged after the action of concentrated hydrochloric acid a t 170° f o r two and a-half hours. The quantity of this compound available was insufficient for the determination of its constitution and we are consequently unable to offer any suggestion as to how one of the carbon atoms of the starting material has been eliminated. It is perhaps worth record-ing that the formula C,R,,0N2 is that of a phenyldihydro-glyoxalone. 5.2; N=13.5; C1=17.2. C1= 17.8 per cent NITRO- ARPLAF;O- AND AIYXNO-GLYOXALINES. 966 Reductim of 5 - R e ? 2 - z e n e a z o - 4 - m ~ t h ~ ~ g ~ ~ o ~ u ~ ~ ~ e m'th Zinc Dust and Acetic A cid.Ten grams of the azoccompound were dissolved in 150 C.C. of boiling 50 per cent,. acetic acid and reduced by adding gradually 16 grams of zinc dust. After removing the zinc as sulphide the liquor was mixed with 20 C.C. of hydrochloric acid evaporated to a syrup and mixed with alcohol when 1.3 grams of ammonium chloride were collected. The alcoholio mother liquor was deprived of the solvent dissolved in water mixed with sodium carbonate, and shaken with ether when 1.6 grams of the base C,,H,,ON,, described below separated as a nearly colourless insoluble crystal-line powder. The ethereal solution left on evaporation 3.3 grams of aniline. From the alkaline liquor 5-5 grams of a mauve varnish were obtained from which only m a l l quantities of crystlalline sub-stances could be isolated by various methods of treatment.The base C10Hl10N3 crystallises from water in small d o u r -less glistening rhomboidal plates which are anhydrous and melt a t 265O (corr.). It is very sparingly soluble in cold water rather more readily in boiling water. Found C= 63.7 ; H = 6.2 ; N = 22.0. ~loBIION (189.2) requires C=63*5; H=5.9; N=22.2 per cent. The hydrochtloride crystallises from absolute alcohol in trans-parent oblong plates which melt a t 206-208° (corr.). It is readily soluble in water concentrated hydrochlorio acid or hot alcohol. The base dissolves slowly in cold 10 per cent. aqueous sodium hydroxide readily on warming and a well-crystallised sodium salt separates from the solution in prismatic needles.This salt is decomposed by carbon dioxide with the regeneration of the base. A solution of the base in aqueous sodium hydroxide gives with Fehling's solution no change iq the cold but a green precipitate on boiling. A solution of the base in nitric acid gives no pre-cipitate with silver nitrate but on the addition of ammonia a white precipitate which dissolves on heating the solution reappears on cooling and is soluble in excess of ammonia. An aqueous solu-tion of potassium permanganate is unaffected by a solution of the base in sulphuric acid but turns green with a solutian of the base in aqueous sodium hydroxide. The base does not couple with sodium diazobeneene-psulphonate in aqueous sodium oarbonate, and when dissolved in hydrochloric acid and mixed with sodium nitrite does not couple with eodium P-naphthoxide.When the hydrochloride is heatled with concentrated hydrochloric Its aqueous solution reacts strongly acid to litmus 256 FARBHER AND PYMAN: acid for two and a-half hours a t 170° it is decomposed with the formation of ammonium chloride and a hydrochloride which crystallises from alcohol in plates melting and decomposing a t about 280° (corr.). 4-Benzeneaeo-2-nte t hytglyoxatine . This was prepared by the action of bonzenediazonium chloride on 2-methylglyoxaline in aqueous sodium carbonate. The crude product readily resinified when boiled with alcohol and only a small proportion was obtained in a pure state. It forms brick-red prisms which melt a t 158O (corr.) and are very readily soluble in alcohol.Found C=64.3; H=5.7; N=30.0. CIOH1,,N4 (186.2) requires C = 64.5 ; H=5.4 ; N= 30.1 per cent. 4-p- Bromob enzen eazo-2-me t hylglyoxaliiz e . This was prepared in good yield by the action of p-brmno-benzenediazonium chloride on 2-methylglyoxaline in aqueous sodium carbonate. It crysballises from absolute alcohol in red, rhomboidal prisms which are anhydrous and melt and decompose a t 200° (corr.). Found N = 21.0. From ordinary alcohol it separates in elongated prisms which lose 2-2 per cent. of water a t 60° in a vacuum. This hydrated form melts a t about 135O when heated quickly and softens a t this temperature when heated slowly finally melting a t about 1900. It can be dehydrate'd by cryst'allisation from absolute alcohol.The reduction of this campound with either stannous chloride or zinc dust and acetic acid led to mixtures of products from which no crystalline compounds except p-bromoaniline and ammonium chloride could be isolated. @,,H,N,Br (265.1) requires N = 21.1 per cent, 2-Phen~l-4-p-7~~~omobenzenea~oglyoxaline. 8.6 Grams of p-bromoaniline were diazotised and the liquor added to 7.2 grams of 2-phenylglyoxaline and 7 0 grams of hydrat?ed sodium carbonate in 4 litres of water a t go the solution being vigorously stirred during the addition Separatim oE an orang NITRO- ARYLAZO- AND AMINO-GLYOXAL~NES. 257 p t d p i t a t e began a t once but was not complete until forty-eight hours had elapsed. The crude product was crystallised from alcobol and gave 13 grams of the pure azo-compound.2-Phenyl-4-p-bromobenzeneazoylyoxoline crystallises from alcohol in clusters of fine orange needles which melt at; 2 0 1 O (corr.) and are anhydrous. Found N = 16.9. C,,H,,N,Br (327.1) requires N = 17.1 per cent. R edzcction of 2-Plzenyl-4-p-b romob enzeneazoglyoxuline with Stawnom Chloride Formation of a Base C,,H,,N4Br. Two grams-of the azorcompound were suspended in 20 C.C. of boiling 5 per cent'. aqueous hydrochloric acid and mixed with 10 C.C. of stannous chloride solution. The solution was Iiltered quickly from a little resinous matter and mixed with 20 C.C. of concentsated hydrochloric acid when a crystalline tin salt separated. This was deprived of tin and the filtrate was evaporated t o a small volume when 0.85 gram of a crystalline hydrochloride separated.This hydrochloride crystallises from dilute hydrochloric acid in nearly colourless needles which after drying in a vacuum melt and decompose a t 255O (corr.). Found C=45*0; H=3.8; N=13.7. C,,Hl3N4Br,2HC1 (402.0) requires C = 44.8 ; H = 3.8 ; N= 13.9 per cent. 0.1530 gave by Carius's method 0.1750 AgCl + AgBr. Calc., It is sparingly soluble in cold water more readily so in hot water. The aqueous solution gradually acquires a purple colour in the air or on the addition of acidified aqueous potassium per-manganate. I n the presence of an exce8s of hydrochloric acid, aqueous solutions are stable in the air. Sodium carbonate or ammonia precipitate the base as a grey flocculent precipitate which is soluble in ether the ethereal solution rapidly assuming a purple colour.On the addition of sodium hydroxide to an aqueous soha tion of the hydrochloride a pale purple solution results. On add-ing sodium diazobenzenepsulphonate to a dilute solution of the compound in the presence of sodium carbonate a dull purple colour is produced. An aqueous solution of the hydrochloride con-taining an excess of hydrochloric acid gives on the addittion of 0.1806 sdium nitrite a deep orange solution which yields a sparingly eoluble purple dye with sodium Fnaphthoxide. On mixing a solu-tion of the hydrochloride in dilute acetic acid with sodim acetate and benzaldehyde there is evidence of the formation of a benzyl-idene derivative. When an aqueous solution of the hydrochloride is mixed with sodium nitroprusside a pale buff precipitate is formed which dissolves in sodium hydroxide giving a deep red soh tioln.The tm'acetyl derivative was obtained by heating the hydro-chloride for one hour on the water-bath with an excess of acetic anhydride and anhydrous sodium acetate. On heating the product with aqueous sodium carbonate it separated as a slate-grey crystal-line powder which did not melt a t 300O. Found C=55*2; H=4.1; N=12-6 12.7; Br=1?.3. C,,H1903N4Br (455.2) requires C-55.4,; H =4.2; N- 12.3; Br=17-6 per cent. It is almost insoluble in boiling water or alcohol and does not dissolve in dilute acids or in aqueous sodium hydroxide. From its mode of forination composition and properties it is dear that the hydrochloride C,,H,3N4Br,2€€C1 arises from 2-phenyl-4-p-br~mobenzenehydrazoglyoxaline by a change of the semidine or benzidine type but it is not possible bo decide ddnitely without further evidence which of the three formulea given below represents its constitution.2-p-Sulphobenaeneasoglyoxalirte-4 5-dicarboxylic A d (XX p. 226). 20.8 Grams of sulphanilic acid were converted into diambenzene p-sulphonic acid and the moist crystals (representing about 20 grams of dry substance) were added tol a cold solution of 16 gram8 of gly~aline-4:5dicarboxylic acid in 240 C.C. of 10 per cent. aqueous sodium hydroxide. After keeping for one and a-half hours the liquor was mixed wikh sufficient glacial acetio acid (36 c.c.) to neutralise tjhe alkali coded and kept for half an hour, when a mass of silky yellow needles-the disodium salt of the ne acid separated.These were recrystalliaed twice from 200 C.C. of wabt and fifinally dissolved in 150 C.C. of hot water afid mixed with 50 C.C. of hydrochloric acid when 12 grams of 2-p-sulphobenzene-azoglyoxaline-4 5-dicarboxylic acid separated in red microscopic prisms mixed with some smaller crystals of glyoxaline-4 5-dicarb-oxylic acid from which it was purified by fractional crystallisation from water. The acid separatm from water with 2H20 which is lost a t 130° in a vaeuum but not a t 100-llOo under normal pressure. Found in air-dried subskance loss at 130° in a vacaum=10*0; C,,H8O,N4S,2H2O (376.2) requires H,O\=9.6 ; C= 35.1 ; H = 3.3; N=14*9; 5=8.5 per cent. It is soluble in aqueous alkalis but not more soluble in dilute aqueous mineral acids than in water.The d’sodizlm salt separates in yellow silky needles which con-tain 3H20 when the acid is dissolved in aqueous sodium hydr-oxide and sufficient acetic acid is added to combine with the alkali. It is readily soluble in hot water somewhat sparingly so in cold. C=35.5; H z 3 . 3 ; N=15.1; S=8.2. It is sparingly soluble in cold water but readily so in hot. Found in air-dried salt loss a t 100°=11-6 12.6. Found in salt dried a t looo S=8*2; X=11.7. C,,B6O7N4SN+~3H20 (438.2) requires 3H20 = 12.3 per cent. CllH,07N,SNa (384.2) requires S = 8.4 ; N = 12.0 per cent. Reduction of 2 - p - SuEphobenzeneasogl?/oxali7e - 4 5 -dz’carboxylic Acid Formatiom of 2-Aminoglyoxaline-4 li-dicmboxylic Acid (XXI p.226). 6-2 Grams of the disdium salt were dissolved in 60 C.C. of 10 per cent. aqueous sodium hydroxide mixed with 12 grams of sodium hyposulphite (80 per cent.) and boiled. The nearly colourless solution was kept overnight acidified with hydrochloric acid, boiled and filtered hot when 1.6 grams of cruds 2-aminolglyvxaline-4 5-dicarboxylic acid separated. This was purified by solution in aqueous sodium hydroxide filtration and reprecipihtion with hydrochloric acid and finally crystallised from about 500 C.C. of dilute hydrochloric acid. 2-A ntinoglyoxaline-4 5-dicarboxyEic acid forms minute pale buff needles which effervesce a t 2 4 5 O (corr.) and then melt. It is very sparingly soluble in cold water a little mom readily in hot 260 NITRO- ARYLAZO- AND AMINO-QLYOXALINES.Found i n substance dried a t l l O o C=34.6; H=3.2; N=24.6. C5H50,N3 (171.1) requires C=35*1; H=3.0; N-24.6 per cent. It is soluble in aqueous alkalis but not appreciably more soluble ia dilub acids than in water. An aqueous solution acidified with sulphuric acid decolorises cold aqueous perrnanganate instantly. When treated with hydrochloric acid and sodium nitrite and poured inta a solution of &naphthol in aqueous sodium hydroxide, it gives a reddish-brown colour . With sodium diazobenzene-p sulphonat,e in aqueous sodium carbonate it gives a reddish-brown colour. It does not give any characteristic colour with sodium nit'roprusside and sodium hydroxide. Action of Water at 170°.-After a preliminary experiment in which it was found that the product contained ammonium carbonate 1-33 grams of the acid and 30 C.C. of water were heated in a sealed tube for twelve hours a t 170° when a dark brown deposit formed. After adding alkali and distilling into standard acid 0-157 gram of ammonia was found whereas 0.132 gram repre-sents the liberation of one molecular proportion. From the resi-due of the distillation small quantities of a crystalline picrate were isolated but in insufficient amount for characterisation. Action of Boiling Aniline.-O.9 Gram of the acid was boiled with 10 C.C. of aniline for six hours under a reflux condenser in which a small quantity of ammonium carbonate collected. The product was distilled wihh steam to remove aniline and left a pale brown aqueous liquor containing some resinous matter. The liquor was cooled filtered and mixed with cold saturated aqueous picric acid when 1.0 gram of a crystalline picrate melting a t about 215O was obtained. After crystallising this from water tlwice it gave 0.4 gram of 2-aminoglyoxaline picrate melting a t 234O (corr.) the pure substance melting a t 236O and a mixture of the two a t 234O in the same bath. From the picrate the hydro-chloride and stannichloride were prepared and identified as the salts of 2-aminoglyoxaline previously described. LONDON E.C.1. THE WELLCOME CHEXICAL RESEARCH LABORATORIES, [Received February 7th 1919.