COLLIE : PRODUCTION OF PYRIDINE DERIVdTIVES, ETC. 299 XXVI.-Production of Pyridine Derivatives from Ethylic p- Amidowotonctte. By J. NORXAN COLLIE, Ph.D., F.R.S. AMONGST compounds from which pyridine derivatives can be obtained, ethylic acetoacetate stands out prominently, and since Hantzsch (Ann., 1882, 215,lO) showed that, by warming this compound with aldehyde- ammonia, die thylic dihydrocollidinecarboxylate was formed, almost every year has seen some addition to the ever-increasing number of300 COLLIE : PRODUCTION OF PYRIDINE DERIVATIVES FROM closed ring nitrogen compounds which can be produced either from ethylic acetoacetate or its condensation derivatives. I have already called attention to the fact that when ethylic /3-amido- crotonate is destructively distilled, various pyridine derivatives are formed (AWL, 1884,226,297; Trans., 1891,59,172 ; also Trans., 1895, 6'7, 2 15), the chief product being ethylic lut3donemonocarboxylate. I have also shown how pyridine derivatives could be produced by the same kind of reaction from dehydracetic acid and from triacetic lactone, both condensation products of ethylic acetoacetate.Some years ago, during attempts to obtain larger quantities of ethylic lutidonemonocarboxylate, I prepared the hydrochloride of ethylic /3-amidocrotonate in the hope that, when it was heated, it would con- dense at a lower temperature, according to the following equation: C6H,,N0,,HC1 + C,H,,NO, = C,oH,3N0, + NH,CI+ C,H,* OH. The re- action took place at once, and a large quantity of a pyridine com- pound was obtained, but I was astonished to find that, not only was the product of the action a mixture of two substances, but that they both had the formula C,,H,,NO,, and also that they were both entirely distinct in properties from ethylic lutidonemonocarboxylate ; their melting points were as follows.Ethylic lutidonemono- Compound A. Compound R. carboxyla te. A short note of the reaction was published in 1887 (Bey., 20, 445), and, since then, from time to time I have investigated the properties of these two substances and their compounds. They are both ethylic salts, and yield alcohol and the sodium salts of acids when warmed with caustic soda solution. 138-139"" 166-167" 163-164" The melting points of the acids are- Lutidonemono- Compound A. Compound B. carboxylic acid. 300-304" 190-1 91 " 254-256" When heated above their melting points, they both decompose quanti- tatively into pseudolutidostyril, a substance first obtained by Hantzsch (Bey., 1884, 17, 2904).Now Iutidonemonocarboxylic acid, when heated, yields lutidone, \CO/ Lutidone. CH3 Pseudolutidostyril. * All the melting and boiling points given have been determined by means of ti set of Auchutz thermometers, and are therefore corrected temperatures.ETHYLIC P-AMIDOCROTONATE, 301 and as both the acids from compounds A and B gave pseudolutido- styril, it seemed possible that the difference between thein might lie in some stereochemical molecular arrangement. After furthur study of their properties, however, I have abandoned this idea, as it is quite possible t o explain their various differences and the produehion of pseudolutidostyril by ordinary graphic formula The interesting fact still remains that, if ethylic P-amidocrotonate is heated alone, it yields lutidone derivatives, whilst its hydrochloride yields two isomeric coni- pounds which are derivatives of pseudolutidostyril. The formation of these two compounds may be explained as follows.C:OOEt* CiHNH,j-C(CH,)=CH -+ COOEt. 8 C(CH,):CH' , ..................... CH,* C-NHiH 1 1 ; ............. :,.. ................... EtO'-~O ! CH; C *NH- 70 i HCI; ! ............. I Etb ylic B-amidocrot ona t e Compound A, in. p. 138-139 hydrochloride. or eth ylic pseudolutidostyril monocarboxylate. whilst the tnutomeric modification of ethylic P-amidocrotonate would give COOEt.CH,*C=NjH I ...., EtO\-- ............... .... . COOEt*CH, Q: N--- 8 OH yo += HC:C( CH,) * CH Compound €3, ni. 1'. 166-167". ............, HCiH,NH'I-C(CH,). CH, ; HC1: Ethylic B-imidobutyrate, ................. This reaction is precisely similar to that which takes place when ethylic 1utidonemonocarboxylate is formed (compare Trans., 1895, 67, 401), when dehydracetic acid is heated with strong ammonia, or when dehy- dracetic acid is produced by the distillation of ethylic acetoacetate. The production of pseudolutidostyril from the free acids of either of the two compounds A. and B is easily accounted for- ,;COO;H-CH,.F '. ....... . _ I Yo \ c HCH ........ ........... HC Yo C 3-1 I ',COO:HC I /NH\ /NH\ ........ CH C. I coo H-cH,*$ CH, Acid A. CH3 Acid 8. These compounds are, therefore, derivatives of isodeh ydracetic acid, /O\ CH,E yo COOHC CH ; VOL.LXXI. Y302 COLLIE : PRODUCTION OF PYRIDINE DERIVATIVES FROM the connection of that acid with pseudolutidostyril was first pointed out by Hantzsch (Ann., 1884, 222, 46). A great deal of the work that has been done with these nitrogen compounds has been with the object of attempting to arrive possibly a t reactions by which it mould be possible to follow, step by step, the formation of true pyridine compounds. These oxy-derivatives are hardly true pyridine derivatives, being almost devoid of basic pro- perties, but yet are as closely allied to pyridine as phloroglucinol is to benzene. When treated with chlorinating agents, they give chloro-deri- vatives which, without doubt, contain the atomic linking present in pyridine itself.One is, therefore, able to follow the gradual change from an open chain compound, such as ethylic P-amidocrotonate, through a series of reactions, none of which need a high temperature, until, finally, true pyridine derivatives are obtained. These oxypyridine compounds, therefore, supply excellent material for the investigation. From their formation and properties, the evidence seems to be more in favour of the linking in the pyridine ring resembling that of benzene, than that it should be represented by the formula suggested first by Lieben and Haitinger. For, although the formation of y-chlorolutidine by treating lutidone with phosphorus pentachloride seems to favour Lieben and Haitinger’s formula, \ C O / Lutidone.\ d d y -Chlorolutidine. still the production of a-chlorolutidine from pseudolutidostyril is in direct opposition to such an atomic grouping, A CH,*S YCl HC CH FH\. + CH,. yo HC CH Pseudolutid OR tyril. a-Chlorolutidine. The explanation which agrees best with all these reactions is that the internal linking of the pyridine ring resembles that of benzene, and therefore it may be expressed either by the original formula for pyridine that Karner first suggested, or by what is practically identical with it, the centric formula,ETHYLIC 6-AMIDOCROTONATE. 303 d ctioit @' Heat on the Hydi*ochlos.icle of' Ethplic P-Amidocrotomte. When perfectly dry hydrogen chloride is passed into a dry ethereal solution of ethylic P-amidocrotonate, a precipitate begins t o form almost at once, and settles to the bottom of the flask as a semi-crystal- line mass.This crystalline compound is at once decomposed by water into ammonium chloride and ethylic acetoacetate, but if collected carefully and analysed, it is found to contain 21.6 per cent. C1, 43.4 per cent. C, and 7.5 per cent, H. C,H,,NO,,HCl requires 22.3 per cent. C1, 43.5 per cent. C, and 7.3 per cent. H. When the crystals are heated they partially melt, and an action a t once occurs with evolution of heat, the mass becoming almost solid from separation of ammonium chloride, By treating the product with a small quantity of water, the ammonium chloride can be washed away, leaving a solid, crystalline residue, which dissolves almost entirely in boiling water ; after filtering from the small quantity of undissolved resinous matter, the filtrate, 'on cooling, sets to an almost solid mass of long, silky needles When purified by several recry stallisations from hot water, the compound melts a t 138-139".The mean of a large number of analyses gave the following result, c. H, N. Found ... . . . . . .. . . . . .. . . . . , . . , 61.6 6.9 7.3 Calculated for C,,H,,NO:~ . . , 61.5 6.6 7-2 The molecular weight was determined by the Rrtoult method, using Found, 194. R1.W. of C,,H,,WO, = 195. The reaction, therefore, occurs according to the equntioii As hydrogen chloride mas evolved during the action, it was thought, that time might be saved by only half-saturating the ethylic P-amido- crotonate with the gas. An experiment was made, using 100 grams of the amidocrotonate dissolved in ether, the solution being divided into two equal portions, one of which was treated with excess of gaseous hydrogen chloride, the other half added, and the ether evaporated off on a water bath.The flask was now transferred to an oil bath and heated t o 120°, when a vigorous reactmion occurred; on cooling, the contents of the flask were treated with a small quantity of water t o remove ammonium chloride, and the residue recrystallised from water, These crystals mere found to be very different from those obtained before; nltimately they proved to be a mixture of two com- pounds, one, A, melting a t 138-139", the other, B, melting a t 166-167" (the chief product). acetic acid as the solvent. BC,H,,NO.,,HCl = C1,,H1,KO;, + CPHR* OH + NH,Cl + HC1.Y 2304 COLLIE : PRODUCTION OF PYRIDINE DERIVATIVES FROM The new compound, B, was anslysed; the mean of several analyses gave, C. H. N. Found . . . , . , . . . . . , . . . . . . . . . . . . . . . 6.8 7.4 Calculated for C,,H,,NO, . . . 61 -5 6.6 7.2 The molecular weight was determined by tbe Raoult method, using glacial acetic acid as the solvent. Found, 195. C,,H,,NO, requires From these analyses, it appears that the compounds A and B are isomeric, and at first it seemed that the change in the conditions of the experiment had determined the formation of the new compound, and that the semi-saturation of the amidocrotonate with hydrogen chloride was the reason for the production of B. This, however, was not found to be the case, for, on attempting to prepare compound A, using crude benzene (b.p. 80-120")instead of ether, and fully saturating with hydrogen chloride, the compound B again seemed to be the chief product. This, however, was not always the case, for in another experi- ment with benzene, A was formed in considerably larger quantities than in the first trial; ultimately, after many different methods had been tried, it was found that, if the solvent used was evaporated off at a low temperature, so as t o leave the hydrochloride in a pure condition, then, on further heating to a temperature of about 120", compound A alone was produced, whereas, if crude benzene was used or excess of amido- crotonate, a varying mixture of A and B was the result. The yield of either of these compounds was never as much as 50 per cent., much resin being always formed at the same time.I n one experi- ment, 400 grams of pure amidocrotonate was saturated with hydrogen chloride in ethereal solution, and 140 grams of A was obtained; whilst, in another experiment, 200 grams that had been semi-saturated with the gas gave 60 grams of B mixed with 5 grams of A. Compound A is easily purified by recrystallisation, but compound B is more difficult to separate from the resin formed a t the same time; of the two, A is also the least soluble in water. Sometimes the separation was effected by warming with dilute solution of soda for a few minutes ; under these conditions, B was a t once hydrolysed, whilst A remained undecomposed, and crystallised out again on cooling the solution.The acid, of which B is the ethylic salt, was recovered by acidifying the soda solution. To sum up,-(1) When 100 grams of ethylic P-amidocrotonate was dissolved in ether, and saturated with dry gaseous hydrogen chloride, and the ether evaporated, the residue, on heating, gave 30 grams of A melting at 138-139". (2) When 100 gram? of ethylic pamido- crotonate was semi-saturated with the gas and heated under similar conditions, 25 grams of €3 was obtained mixed with small quantities of A ; the yield of B, however, often varied in different experiments, 61.8 M.W. = 195.ETHYLIC P-AMIDOCROTONATE. 305 The theoretical amount that ought to be produced is '71 grams, the yield, therefore, is only 40 per cent. Of the various experiments tried, it may be mentioned that B was also produced in small quantities when ethylic P-amidocrotonate was treated with acetyl chloride.Both A and B are ethereal salts of acids, and yield alcohol and a sodium salt when boiled with caustic soda. But A has to be per- sistently boiled with strong soda solution before the decomposition is complete, whilst B, as has been already mentioned, is hydrolysed at once. Many attempts were made to convert the one into the other, but boiling with acids, heating with water alone in sealed tubes, or with acid did not effect the change. The compound A crystallises from hot water in long silky needles, but these become granular by stirring or standing when it is quite pure, and settle down to a thin layer at the bottom of the crystallising vessel. When boiled with acetyl chloride, no change occurs, and the substance may be heated with acetic anhydride at 140" for 6 hours with no result.Hydroxylamine does not seem t o form any compound, and nitrous acid or boiling hydrochloric acid is without action on it. If it is heated a t 100" with strong sulphuric acid and the mixture then poured into water, the unchanged compound separates out again. The action of heated soda lime was also tried. 10 grams was mixed with and distilled over excess of heated soda lime; traces of ammonia were produced, but the chief product was an oil boiling at 280-285". This gave a well crystallised platinochloride, containing no water of crystallisation ; the mean of four analyses gave 25.9 per cent, Pt. When the compound A is dissolved in glacial acetic acid, and bromine is added in excess, a bromine compound is produced which can be preci- pitated by pouring the mixture into water; it recrystallises from alcohol in long, glistening needles, melts a t 158-159', and on analysis proved t9 be a monobromo-substitution product.The mean of several analyses gave C = 44.0 ; H = 4.5 ; N = 5.3 ; Br = 29.1. Calc. for CioHl,NO,Br : C = 43.8 ; H = 4.4 ; N = 5-1 ; Br = 29.2. Compound A also reacts with pentachloride of phosphorus. It is best not to dilute the substance with any solvent, but to add the penta- chloride in small quantities, keeping the flask heated in an oil bath a t 180". Theory for (C,HI,NO),,H,PtC1 is 26.2 per cent, The reaction is as follows :- CioH1,NO, + PCI, = C1,H12N02C1 + POCI, + HCI. The action is not violent, and after the oxychloride of phosphorus has been distilled off, the residue can be added to water and steam distilled.The oil which passes over when purified boils at 288-290" Its analyses gave the following numbers, C = 56.6; H = 6.0306 CO1,LIE : PRODUCTION OF PYRIDINE DERIVATIVES z;"I:OX N = 7.0 ; C1 = 16%. Calc. for CI,H,,NO,C1 : C: = 56.2 ; H = 5.6 ; The chloride is a very stable substance, and is only decomposed slowly by boiling with potash. Some of it was treated for a week with t i n and strong hydrochloric acid warmed on a water bath. By blowing steam through the neutralised product of the reaction, an oil boiling at 246-5348' was obtained, which, unlike the original chloride, had basic properties and formed a platinum salt melting a t 208-210".When this oil is boiled with aqueous soda and an acid added to the solution, a pyridine acid melting a t 158-160" is precipitated. The platinum salt contained 25.4 per cent. of Pt, and the ether and acid were both free from chlorine. The substance which had been produced was, therefore, an ethplic ay-dimethyl pyridine /3-carboxylate, and is probably identical with a compound produced by Michoel (Re?*., 1586, 18, 2020) from ethylic aceqo- acetate. He found that the ether, C5H,N(CKJ3* COOC,H,, boiled a t 246-24'i0, the platinochloride melted a t l o l o , and the acid had a melting point of 166". This reduction of the chloro-derivative by tin and hydrochloric acid is of interest, because many of the chloropyridines are not acted on by nascent hydrogen produced in this manner, The compound A needs prolonged treatment with boiling and moderately concentrated soda t o effect its decomposition.The sodium salt produced is decomposed on the addition of hydrochloric acid, and the free acid is a t once precipitated. It is very insoluble in most solvents, but can be recrystallised from boiling water. When pure, it melts at a little above 300", about 304', and decomposes a t once at that temperature, carbon dioxide and pseudolutidostyril being produced. It crystallises with 1 mol. H,O, and has the formula C,H,NO,,H,O. Auchutz, Bendix, and Kerp (Ann., 1890,259, p. 174) obtained an acid melting at 275" by heating isodehydracetic acid with ammonia ; although they give the melting point as 2 7 5 O , it probably is the same acid, for, on heating, it yields carbon dioxide and pseudolutidostyril. The silver salt is thrown down as a white precipitate from neutral solutions.The coppel' salt is a light green precipitate. The Zecccl salt crystallises from concentrated solutions in small tufts composed of microscopical needles. The bawhra salt is also soluble, but crystallises from concen- trat ed solution e. Several analyses of the acids mere made which gave results agreeing with the formula C,H9N0,,K,0. As already st,ated, when it is heated it decomposes at its melting point, giving carbon dioxide and pseudolutidostyril N = 6.5; C1 = 16.6. Several salts of this acid were prepared and annlysed.ETHYLIC P-ARZIDOCROTONATE. 307 This reaction appears to be nearly quantitative. Five grams of the dried acid yielded 660 C.C.of carbon dioxide, and 3.6 grams of residue, the amount required by theory being 670 C.C. and 3.7 grams. The residue agreed in every respect with pseudolutidostyril ; it melted a t 180--181", and boiled at 306-301". The analyses t h a t were made also agreed with the formula C,K,NO. This substance was first prepared by Hantzsch (Bey., 1884, 17, 1026) by the action of sulphuric acid on CH, OH ( p n \(g \(+ CH, yo CH,*COOH, /?.< /"\ CH,<* $! I=;.CH, CH3. COOH. C C-COOH = HC CH + co, I I CH.3 A l e thylpseudolutidostyril. The methylpseudolutidostyril was then converted into pseudolutido- styril by heating with hydriodic acid. As very considerable mole- cnlar change must have occurred during the production of methylpseudo- lutidostyril, it is of interest to be able t o confirm the formula which Hantzsch suggested for pseudolutidostyril from the formation of that substance by a totally different set of reactions, CH./NH\ Yo CH Pseucloluticlosty Ti2 3 E HC According to Hantzsch, pseudolutidostyril, when heated with zinc dust, gave ny-dimethylpyridine. Before I had seen Hantzsch's paper, I had tried the experiment, and as the results were slightly different, and the amounts used very much larger, they are worth recording. I n one experiment, where 25 grams of the substance was heated with zinc dust, 12 grams of pyridine bases were obtained boiling between 150 and 170"; these, on fractional distillation, gave, at 150-160", 2 grams; 160--165", 6 grams ; 165-170", 4 grams. The portion boiling from 150-160" was converted into platminochloride and re- crystallised; it contained 31.0 per cent.Pt, and mas without doubt the substance [C,H,N(CH,),],,H,PtC1,, which contains 3 1 *2 per cent. Pt and has no water of crystallisation. ay-Dimethplpyridine boils at, 166-157".308 COLLIE : PRODUCTION OF PYRIDINE DERIVATIVES FRON The larger portion of the pyridine bases, however, boiled between 160" and 170°, and gave a platinochloride less soluble in water than the former one. After many crystallisations, the salt, which melted at 217", was repeatedly analysed. (1) C = 29.4 ; H = 3.9. (2) C = 29.5 ; H = 4.0. (3) C1= 33.1. (4) Six determinations of platinum, varying from 29.7 t o 29.9. The substance is, therefore, a trimethylpyridine or a collidine.CCSH~N(CH&,, H,PtCI, * [C,H,N( CH3)&H,PtClp Found. Lntidine. Collidine. C 26.9 29-4 29.4 H 3.2 3.7 3.9 C1 34.1 32.7 33a1 Pt 31.2 29.9 29.9 Some of the platinochloride was decomposed with hydrogen sulphide, and the free base obtained from the filtrate; its boiling point was 167-168". It gave C=79*2; H=9.1 ; N (by diff.)=11*7. C. H. N. Lutidine, C7H,N.. ................ 78.5 8.4 13.1 Clollidine, C,H,,N ............... 79.3 9-1 11.6 Found .............................. 79.2 9.1 11.7 From the boiling point, probably, this collidine is the symmetrical trime t hy lpyridine. According to Hantzsch (Ann., 1882, 215, 13), this collidine boils at 171-172", whilst Durkopf (Ber., 1888, 21, 2713) gives the boiling point 167-168". The remainder of the base was submitted to oxidation by perman- ganate of potash-for, according to Durkopf, uvitonic acid is pro- duced (pyridine-a-methyl-a'y-dicarboxylic acid), but the products were formed in such small quantity that no definite results were obtained.That it is a trimethylpyridine is of considerable interest ; and it is a curious fact that a substance like pseudolutidostyril should be thus changed into a collidine derivative when heated with zinc dust, and that Hantzsch's collidine compound with strong sulphuric acid should yield pseudolutidostyril. Moreover, the chlorolutidine obtained by the action of pentachloride of phosphorus on pseudolutidostyril seemsETHYL~C P-AMIDOCROTONATE. 309 to give nothing but lutidine (b. p. 156-158") when heated with zinc dust. This chlorolutidine is produced almost quantitatively, 25 grains of pseudolutidostyril yielding about 22 grams of the chloro-compound, a-Cl~loi*o-a'y-c~irnetl~l2~?/~idi7ze, C5NH2(C€33)2Cl.When pure, it boils at 212-214". An analysis gave C=59*0, H = 5.7, whiIst C7H,NC1 requires c' = 59.3, It combines feebly with hydrochloric acid, and forms a stable platinochloride, which, when analysed, gave 28.1 per cent. Pt. Theory for (C7H8NCl),,H,PtCl, = 28.1 per cent. The only may to prepare the hydrochloride is to saturate the oil with hydrogen chloride. The white solid hydrochloride thus formed was weighed, and it was found that the chlorolutidine had taken up 12.2 per cent. of hydrogenchloride, which is almost the amount necessary if 2 molecules of the chloro- lntidine react with 1 molecule of hydrogen chloride-1 2.8 per cent.H = 5.6. 2C7H,NCl + HC1= ( C7H,NCl),,HC1. The platinochloride seems, however, t o be the normal compound, for it gave 28.0 per cent. Pt, and (C7H,NC1),,H2PtCl, contains 28.1 per cent. Pt. /NH\ \ C / CH, Yo COUZ~OU?.L~ B, m. p . 166-167", COOC2H5* CH,* HC CH I The analysis and methods for the preparation of this substance have already been given (p. 304). When it is boiled with strong hydrochloric acid, it slowly decom- poses into the free acid (m. p. 190-191"), carbon dioxide, ethylic chloride, and pseudolutidostyril. It does not give an acetyl derivative when heated with acetic chloride or acetic anhydride, but, as already stated, it is hydrolgsed at once by alkalis (diffarence from compound A), and on adding an acid to the alkaline solution, the free acid crys- tallises out.Unlike the compound A, it does not react in a simple manner with pentachloride of phosphorus. Even when it is dissolved in various solvents, and the pentachloride of phosphorus is added carefully, con- siderable decomposition occurs, blackening, evolution of carbon dioxide and other gases being amongst the most noticeable results, whilst no pure substance could be obtained from the product of the action. With bromine, however, it forms a definite compound. When the compound310 COLLIE : PRODUCTION OF PTRIDINE DERIVATIVES, ETC. B is dissolved in glacial acetic acid, and bromine is added carefully, two molecular proportions of bromine have to be added before it re- mains in excess. The mean of three bromine determinations gave 44.7 per cent.Br., and on combustion C = 34.5 per cent., H = 3.3 per cent. The pure bromine compound melts a t 168-170". C10HllN03Br2 requires C = 34.0 ; H = 3.1 ; Br = 45.3 per cent. The bromide is hydrolysed a t once on treatment with soda; the free acid crystallises in small needles, and when heated to 256-258", melts and chars, giving off carbon dioxide, bromacetone, and hydrogen brom- ide. Although the compound A gives no hydrazone with phenyl- hydrazine, the compound B, when heated in a sealed tube with that re- agent, enters a t once into combination with it. It was found, however, that, even after many crystallieations, the substance was not pure. Results of analysis : C = 65.0 ; H = 6.3 ; N = 15.0 per cent. Calculated C,GH,,N,O, : C = 67.4 ; H = 6.6 ; N = 14.7.This compound, when heated, melted a t 227-228", blackening con- siderably, and gave a strong smell of aniline compounds, so that, although it had not been obtained in a pure condition, still there was no doubt that a hydrazone had been produced. When the compound B is treated even with dilute alkalis, i t dissolves, and on warming a t once hydrolyses. The free acid, which is obtained by adding hydrochloric acid to the solution of the sodium salt, crys- tallises from water, either in needles or flat, needle-shaped plates, con- taining no water of crystallisation. It melts a t 190-191", and on analysis gave results agreeing with the formula C,H,NO,. The sodium salt crystallises in large plates ; the bnrium salt seems t o be less soluble in boiling than in cold water ; the lead salt separates in leaflets from strong solutions; the silvw salt is thrown down as a flocculent precipitate in neutral solutions, and is moderately soluble in hot water, from which it can he crystallised. It contains no water of crystallisation, and gave on analysis 56.0 per cent. Ag. Theory for C,H,NO,Ag, : Ag = 56-0 per cent. Probably the hydrogen atom that is attached to the nitrogen atom, has beenreplaced by the silver as well as the hydrogen in the carboxyl group. The same result was obtained on analysing the silver salt of the isomeric acid (m. p. 300-304") from the compound A. 54.1 and 55.0 per cent. Ag. On heating this acid (m. p. 190-191"), it is noticed that copious effervescence of carbon dioxide occiirs the moment it melts; 10 grams lost 2.5 grams. According to the equation, C,H9N0,, = C,H,NO + CO,, it should lose 26 per cent. The residue in the flask is almost pure pseudolutidostyril, identical in ,,YOUNG : OXIDATION OF PHENYLSTYRENYLOXYTRIAZOZE. 311 every respect with that obtained by heating the acid (m. p. 300-304"). The simplest method of explanation is that the acid has the following formula : /NH, / \ HC CH ff.200 H*CH2*E Yo That it is a substituted acetic acid also receives additional support from its behaviour with ferric chloride. I n neutral solutions, it gives at first a reddish coloration, bnt on boiling, the yellow ferric salt is at once precipitated,