238 REVIS AND KIPPING : DERIVATIVES OF a-HYDRINDONE. XX.-Des*ivativcs of a-Hydridone. By CECIL REVIS, Assoc. C.G.Inst., and FREDERIC STANLEY KIPPINB, Ph.D., D.Sc. ALTHOUGH a-hydrindone has now been known for some years, and its preparation by the method previously described by one of us (Kipping, Trans., 1894, 66, 4SO) is a comparatively simple matter, its derivatives have not yet received much attention, and little is known of their general chemical behaviour ; for this reason, we have taken up again the study of hydrindone derivatives where it was left some years ago. I n the course of this work, we have not only prepared and character- ised a number of new derivatives, but we have also examined more closely the properties of some of the compounds previously described, more particularly in order to compare their behaviour with that of corresponding derivatives of camphor.Between a-hydrindone and camphor there is, in some respects, a con- siderable difference in constitution, but, at the same time, the two ketones are doubtless of a somewhat similar structure as each consists of two closed chains, certain carbon atoms of which are common to both. Expressing the constitution of camphor by Bredt’s formula, each compound contains a 5-carbon nucleus in which the group -CK,*CO- is present; in other words, both a-hydrindone and camphor contain tho complex H I I I - C -CH, -c- I I -c --co I upon which is built a second closed chain. Now, since all the more important reactions o r changes which these two compounds undergo are determined in a greater or less degree by the presence of the -CH,*CO- group, it might be expected that there would be an obvious and close similarity in the behaviour of the two ketones, especially as regards those reactions in which this group is known to be concerned.Investigation has shown that this is so, but only to a very limited exteut : for instance, just as camphor and a-monobromocamphor are oxidised by nitric acid, giving camphoric acid, C,H,,(COOH),, so, like- wise, hydrindone arid rnonobromhyclrindone are converted into phthalic acid, C,H,(COOH), under similar conditions. Again, dibromhydrin- done, like a-dibromocamphor is a very stable substance, and may beREVIS AND KIPPING : DERIVATIVES on ~-EIYDRINDONE. 239 crystallised unchanged from boiling nitric acid of sp. gr.1.4; in both cases, therefore, the substitution of bromine for the 2 atoms of hydrogen of the -CH,* CO- group confers great stability towards nitric acid. The formation of an isonitroso. derivative, containing the complex -C(NOH)*CO- is another point in which the two ketones resemble one another, but, without going into details, it may be stated that here any close analogy ceases to be apparent. One property stands out very prominently in the case of hydrindone, and that is the readiness with which this ketone forms condensation products, such as anhydrobishydrindone, truxene, benzylidenehydrin- done, acetonehydrindone, &c. (Zoc. cit.) ; camphor, on the other hand, shows no marked tendency either to undergo condensation with itself or to interact with aldehydes and ketones; although when treated with energetic dehydrating agents, such as phosphorus pentoxide, zinc chloride, &c., i t loses the elements of water giving cymene, or loses hydrogen giving carvacrol, it yields no compounds analogous to anhy- drobishydrindone or truxene, and although, in the form of its sodium derivative, it interacts with a number of aldehydes under particular conditions, as has been shown by Haller (Conzpt.rend., 1891, 113, 22), yet it does not seem t o do so under conditions which, in the case of hydrindone, at once cause the formation of a condensation product. This last fact is all the more remarkable, inasmuch as Wallacb has lately shown (Bey., 1896, 29, 1595) that a great number of cycloid ketones of various kinds interact readily with benzaldehyde, giving mono- and in some cases di-benzylidene derivatives.We have also found that the mono- and di-bromo-derivatives of a-hydrindone behave very differently from the corresponding deriva- tives of camphor when submitted to the action of alcoholic potash. MThereas a-monobromocamphor is very stable, and even when attacked is simply reduced t o camphor, monobromhydrindone is very readily acted on by alcoholic potash at ordinary temperatures, giving a sub- stance of the composition C18Hl:3Br02, two molecules of the monobromo- compound uniting together whilst the elements of hydrogen bromide are separated. This condensation product, which we name hycll.inclonyZbl.onz~.yd~win- done, and which has probably the constitution represented by the C,H aCH, CH,-C H formula 1 ' I I I ' ' crystallises in transparent monosym- CO--CH-- CBr -CO metric prisms and has been examined crystallographically.Dibromhydrindone, unlike a-dibromocamphor-which is simply con- verted into monobromocamphor-also gives a condensation product when treated with alcoholic potash ; this compound has the composition CISH1lErOS, so that its formation may be expressed by the equation240 REVIS AND KIPPING : DERIVATIVES OF a-HYDRINDONE. 2C,H,Br,O - HBr - Br, = C18H,,Br0,. It seems probable, therefore, that it is an ~~ndonyZbroml~?/d?.i.ndon.e of the constitution represented C H *CH, CH*Y,H, CO--CBr-C--CO by the formula l 6 I I1 This substance is interesting on account of the eagerness with which it takes up benzene, forming with it a compound of the composition ClsHl,BrO,,C,HG, which crystallises in long, flat prisms.A somewhat similar condensation product is obtained when dibrom hydrindone is treated with sodium ethoxide in alcoholic solution ; this substance seems to have the composition Cl,HloBrO,*OEt, but being of little interest, it was not examined very fully. The tendency of hydrindone to undergo condensation is brought out again by the fact that, on treating the ketone with a warm solution of bromine in excess of caustic soda, a sparingly soluble substance of the composition Cl,H1,O, is finally formed. I n some experiments, we also observed the formation of dibromhydrindone, which seemed to indicate that this might be an intermediate product in the conversion of the ketone into the condensation product', but attempts to prepare the latter directly from dibromhydrindone mere unsuccessful.The consti- tution of this condensation product is, perhaps, represented by the C,H,*CH, CH* C H formula I I II I * and for the sake of reference we name CO--C(OH)-C- CO the substance i ? z d o n ~ Z ~ 2 / d r o : 7 ~ ~ 2 / ~ ~ ~ ~ ~ z ~ ~ ~ ~ ; but in this, as in the fore- going cases, the evidence is too slight to afford a sure basis for a con- stitutional formula, The well-known conversion of camphoroxime into canipholenic nitrile being a reaction of great interest, we made experiments with hydrin- done-oxime to see whether the latter could be caused to undergo a similar change, or whether it could be converted into hydrocarbostyril by treatment with mineral acids, just as i t had been previously found to be transformed into that substance when submitted to the action of phosphorus pentachloride (Zoc.cit.). Unfortunately, these experiments gave results of little interest ; the various agents which were tried apparently resolved the oxime into the ketone, the only products being anhydrobishydrindone and truxene, both of which are produced by the condensation of the ketone. An interesting sodium derivative of isonitrosohydrindone was dis- covered in the course of our work; this crystallises from wet ether in pale yellow prisms, which, when gently heated, become bright scarlet, owing to a change in crystalline form; on keeping the scarlet modifica- tion a t the ordinary temperature, it passes spontaneously into the pale yellow form in the course of a few hours. The corresponding potassium derivative has similar properties, and both compounds are decomposed,REVIS AND KIPPING : DERIVATIVES OF WHYDRINDONE.241 to a greater or less extent, into isonitrosohydrindone and alkali hydr- oxide, when attempts are made to recrystallise them from hot water. The hydrocarbon, hydrindene, has been prepared from coal-tar indene by Krfmer and Spilker (Ber., 1890, 23, 3276) and has also been obtained synthetically by Perkin and Revny (Trans., 1894, 66, a%), but owing to the difficulties met with in both cases, our knowledge of the properties of this hydrocarbon is still incoinpIete. In the hope of obtaining it from hydrindone, we repeated and extended K6nig’s work (Insug.Diss., Leipzig, 1889), trying first to reduce the ketone directly to the corresponding alcohol, and afterwards to prepare the alcohol indirectly by reducing hydrindone-oxime and then treating the base thus obtained with nitrous acid. We can fully confirm Kijnig’s con- clusion that, owing to the formation of resinous products in most of the necessary operations, i t is impracticable to use hydrindone as n starting-point for the preparation of hydrindene. The conversion of hydrindone-oxime into amidohydrindone, however, takes place practi- cally quantitatively, and we have thus been able to further characterise the base by preparing and examining a number of its derivatives. The oxalate, acetate, and nitrate are described and also its benzoyl and benzylidene derivatives. EX PER IMENT AL.The first attempts which were made to prepare hydrindone by the method which had been so successfully employed only a year or so before (Trans., 1896, 66, 484), gave very poor results, and although particular care was taken to adhere strictly to the conditions which had previously been found to ensure success, the yield of hydrindone was very small ; i t was thus ascertained that the quality of the aluminium chloride is also a most important factor in determining the course of the action, and that if this substance has been kept for some time even in well-stoppered bottles, the slight decomposition which it undergoes through absorption of moisture renders it practically useless for the conversion of hydrocinnamic chloride into hydrindone.Freshly-pre- pared aluminium chloride, however, gave results as satisfactory as those previously recorded. Hyd~*indonesemicar6ccxoute, C,H,:N*NH* CO *NH,. This is very easily prepared by dissolving the ketone in dilute alcohol, adding solutions of semicarbazide hydrochloride and potassium acetate in dilute alcohol, and then warming on the water bath. After a short time, a white, crystalline powder begins to separate, and finally the whole solution becomes a pasty mass of crystals. The product is collected, washed with water, and recrystallised from dilute acetic acid242 REVIS AEU'D KIPPING : DERIVATIVES OF a-HYDRINDONE. from which it separates in beautiful, transparent prisms. These crys- tals are hydrated, and seem to contain 7H,O; for the combustion, a sampIe was dried at looo nntil constant in weight.0.1686 lost 0,0654 II,O a t 100'. 0 2692 lost 0,1074 II.,O at 100. 0,2265 gave 0.5256 GO, and 0,1216 H,O. H,O- 38.8. I-I,0=39*9. C = 63.3 ; H = 6.0. C',,,HllN,,O + 7H,O requires H,O = 40.0 per cent. C,,H,,N,O requires C = 6 3 5 ; H = 5.0 per cent. Hydrindonesemicarbazone is only sparingly soluble in dilute alcohol, from which it is deposited in the form of a crystalline powder; it is also rather sparingly soluble in ethylic acetate, and insoluble, or nearly so, in chloroform, benzene, and light petroleum. The hydrated crystals effloresce on exposure to the air, and rapidly lose their water of crys- tallisation when kept over sulphuric acid, crumbling to a colourless powder ; when heated moderately quickly, the anhydrous substance turns brown a t about 220°, melting and decomposing a t about 239".Oxidat ion of IIyclr.i.izdo.ize and of Il.loizobrontl~~cli.i.izdone with Nitric Acid. Kijnig has already shown that hydrindone is oxidised to phthalic acid by dilute nitric acid, and our experiments gave a similar result; homophthalic acid, even if formed at all as an intermediate product, is not present in any appreciable quantity, after the solution has been evaporated to crystallisation ; small quantities of a crystalline product, insoluble in a solution of sodium carbonate, were obtained in some experiments, but not sufficient for analysis. Monobromhydrindone (Trans., 1804, 66, 500) is also easily oxidised by boiling 10-20 per cent. nitric acid; in this case, phthalic acid is produced, together with considerable quantities of dibromhydrindone.It seemed probable a t first that the dibromo-compound was present as impurity in the sample of monobromhydrindone which we used, the former being invariably produced in small quantities on treating hydrin- done with one molecular proportion of bromine ; but as samples of the monobromo-derivative, which had been purified by distillation in steam (dibromhydrindone volntilises very slowly), and by recrystallisation, in- variably afforded the dibromo-compound on oxidation with nitric acid, we are convinced that this is due to an action analogous to that ob- served in the oxidation of a-bromocamphorsulphonic acid. As has been recently shown (Lapworth and Kiyping, this vol., p. l), this monobromosulphonic acid yields, amongs t other products, a considerable quantity of di-bromocamphorsulpholactone when it is boiled with nitric acid, the bromine liberated from that portion of the acid which undergoes oxidation to sulphocamphoric acid acting on some of the unchanged compound.I n the case of monobromhydrindone, there-REVLS AND KIPPING : DERIVATIVES OF U-HYDRINDONE. 243 fore, it mould appear that a portion is oxidised to phthalic acid with liberation of bromine, which then converts the still unchanged mono- bromo-compound into dibromhydrindone. Dibromhydrindone may be crystallised from hot nitric acid (sp. gr. 1-4), by which it does not seem to be oxidised even when the solution is boiled for some time; it is also attacked with dif-liculty by a boiling solution of potassium permanganate in acetic acid.It dissolves in hot, concentrated sulphuric acid, but is precipitated unchanged on diluting, even after the solution has been gently heated ; quinoline a t 100' seems to be without action, although at higher temperatures tarry matter is produced. These facts show that dibromhydrindone is a relntively stable substance, comparable to dibromocamphor, except that it is easily attacked by alcoholic potash, as is shown later. H y d h d o n ylbrom?u&GwZo ne, C, ,Hl 3Br 0,. When a solution of alcoholic potash is added to a cold alcoholic solution of monobromhydrindone, a beautiful violet coloration is ob- served, due possibly to the formation of an alkali derivative, somewhat analogous to that of methyl-ay-diketohydrindene (Wislicenus and Kotzle, Anncclen, 1889, 252, 80) ; on continuing the addition of the alkaline solution, the violet gives place to a red or crimson colour, and potassium bromide is precipitated.As soon as the solution acquires a faint alkaline reaction, it is diluted with water, whereupon a grey, flocculent substance is precipitated, the potassium bromide passing into solution ; this precipitate is separated, washed with water, dried, and recrystal- lised two or three times from hot chloroform, from which it separates in colourless prisms. A sample which had been kept over sulphuric acid was analysed, with the following results. 0.1532 gave 0.3564 CO, and 0.0608 H,O. 0.2212 gave 0.1220 AgBr. Br = 23.47. C,,H,,BrO, requires C = 63 *34. This substance, therefore, is a condensation product, two molecules of the monobromhydrindone having combined with separation of hydrogen bromide ; the unsaturated ketone, indone, the formation of which might have been expected, does not seem to be produced under the above conditions.Hydrindonyl bromhydrindone crystallises from hot chloroform, benzene, or acetic acid, in which it is readily aoluble, and from hot methylic alcohol, in which it is only very sparingly soluble, in small, lustrous, transparent needles or prisms, but on allo\v- ing its solutiolls in cold chloroform or ethylic acetate to evaporate spontaneously, it is deposited in massive crystals (see below). Its melting point is very indefinite, partly, doubtless, because the substance decomposes, partly, perhaps, owing t o the existence of different crystal- C = 63.45 ; H= 4.41. H = 3.81; Er = 23.46 per cent.244 REVIS AND KIPPIKG : DERIVATIVES OF a-HYDRINDONE. line modifications ; whereas, for example, large prisms obtained from chloroform or ethyIic acetate solution melt gradually from 170" t o 178' turning black, similar prisms, previously crushed to a powder, melt sharply a t about 166' when quickly heated from about 130', at which latter temperature the large crystals appear to crack and undergo some change in form; sometimes, however, the crushed prisms melt gradually from 166' to 175", whereas the minute needles deposited from hot, dilute acetic acid melt gradually between 158' and 189', decomposition taking place in both cases.When warmed with a solution of phenylhydrazine in acetic acid, hydrindonylbromhydrindone gives a deep blood-red coloration, but a crystalline hydrazone could not be isolated, the product being a tarry mass ; resinous or tarry substances are also formed when the compound is boiled with alcoholic potash.For the following description of the crystallographic properties of this substance, we are indebted to Mr. Davis, Assoc. C.G.Inst. System. --Monosynz;nzetvio. Fo~ms o6sevuecl. a = (100) c = (001) = ( i o i ) r = (101) p = (110) o = ( 2 2 i ) Angle. ap = 100 :I10 ca = OOL : 100 T'C = 101 :001 pp = 110 :110 cr = 001 :LO1 YT' = 101 : 1 0 1 ar = 100 :lo& t o = 001:22& po = 110:221 cp = 001 :110 ar' = 100 : 101 No. of observations. 24 14 26 14 5 5 1 2 4 4 3 4 Limits. 59'41'-60'59' 59"2' -59'30' 83'30'-85'3' 32'50'-34'0' 30'1 1'-30"35' 6 3"l 2'-64"3 5' 61'31'-62'1 2' 53"9' -54"5' 70"2' -71"9' 21'13'-22"40' 92"2' -93'0' Mean observations.60'22' 59'16' 84'23' 33"38' 30"23' 63"57' 61"53' 53"33' 70'49' 21'58' 92'32/ Calculated. - 59"16' - - 30"27' 64"5' 61"59' 53"56' 70"54' 21"52' 92'46'REVIS AND I~IPPING : DERIVATIVES OF ~-NYDRINDONE. 245 The crystals submitted were eminently unsuitable for goniometrical measurement. Though most of the faces were well developed, hardly one gave a distinct reflection. Many of the faces were so corroded that they gave no images a t all ; others, having a serrated surface, gave long strings of images G f equal brightness. A large number of crystals were measured, but only one or two zones of measurement could be obtained from the same crystal, the corrosion being very irregular.The faces of the form 7’’ (101) appear only as thin strips, both faces of the form not always being present. The faces of ~(101) are also very small, and give bad reflections. The pinakoid ccf100) is the best developed form, but its faces are always very much corroded. There is a fair cleavage parallel to ~(001). The extinction on ~ ( 1 0 0 ) is straight. The interference figure is visible through a cleavage plate parallel to c{OOl), the double refraction being positive. The optic axial plane is nearly parallel t o n(100). The optic axial angle is large. I f ~ d o n yZb.,.onz?~.yd.,.indo./ze, C,,H11Br02. In preparing this substance, dibromhydrindone (5 grams) is dissolved in boiling alcohol (about 25 c.c.), and the solution is then rapidly cooled, so that the substance which separates is obtained in a fine state of division; a solution of potassium hydroxide (about 2.5 grams) in aqueous alcohol is then slowly added, keeping the mixture cool, until t>he latter acquires a permanently alkaline reaction.A t first, the mixture becomes a dirty green, darkening, as the alkali is added, to an indigo, and finally to a deep brown colour, during which changes the dibromhydrindone passes gradually into solution ; potassium bromide is then deposited, together wit’h long, colourless needles of the new substance. The latter is separated, mashed with warm water, dried on porous earthenware, and then boiled with a little light petroleum, t o dissolve out any unchanged dibromhydrindone ; the residue, on being re-crystallised from boiling benzene, separates in beautiful, coloar- less needles or flat prisms.Further, small quantities of this product may be obtained by evaporating the original alcoholic filtrate and then proceeding as befoTe, but this portion is less easily purified. Analyses of a preparation which had been crystallised from benzene and then kept over sulphuric acid for some time gave results which did not agree with those required for any probable formula, and the combustions made a t different times did not give concordant results; this was due to the fact that the substance crystallises with 1 mole- cule of benzene, and although the latter is rapidIy expelled at looo, i t seems to escape only very slowly in a desiccator at the ordinary temperature. A sample, freshly crjstallised from benzene and then exposed for a short time in the air, was annlysed with the following results.VOL. LXXI. S246 REVIS AND KIPPING : DERIVATIVES OF a-HTDRINDONE. 0.2156 lost 0 0457 a t looo, darkening slightly. 0.i468 gave 0.3680 CO, and 0-0540 H,O. C',H,, = 21.2. C = 69-05 ; H = 4.1. C,,H,,BrO,, C,H, requires C = 69.0 ; H = 4.1 ; C,X6 = 15.7 per cent. Samples recrystallised from chloroform, however, gave the following results :- 0.1525 gave 0,3584 CO,, and 0.0445 H,O. C = 64.1 ; 13 = 3.2. 0.2336 gave 0.5465 CO,,and 0.0688 H,O. c1 = 63.8; H = 3.3. CISHIIBrO,, requires C: = 63.7; H = 3.2; Br = 23.6 per cent. Two determinations of the bromine gave 23.5 and 23.3 per cent. respectively, but there is some doubt as to whether the samples were obtained by crystallisation from ethylic acetate, or whether they were crgstallised from benzene, and then left over sulphuric acid f o r some months ; in any case, the abov6 analyses are sufficient to establish the composition of the substance.The crystals containing benzene are long, colourless, flat prisms which, when slowly heated from about SO", sinter and begin to turn brown at about 110"; as the temperature rises, the substance becomes solid again and lighter coloured, until, at about 150", i t decomposes and blackens, giving off gas ; when, however, the capillary tube containing the crystals is plunged into the bath, previously heated to 130", the substance melts immediately, and effervesces, turning reddish-brown, then solidifies, and melts again at about 150°to a black liquid.As already stated, these crystals seem t o lose their benzene, but only very slowly, when they are kept over sulphuric acid, as they undergo slight discoloration, and also change in form, becoming a mass of small needles ; when warmed with methylic alcohol, they become opaque, and change into a white powder consisting of microscopic needles. If, on the other hand, the substance is quickly crystallised from acetic acid, i t is obtained as a colourless powder, which, when covered with warm benzene, rapidly increases in bulk and changes to a mass of long prisms owing to combination with the solvent. Indonylbromhydrindone is very readily soluble in chloroform, but only moderately so in acetic acid, and very sparingly in cold rnethylic alcohol; it separates from a cold mixture of chloroform and methylic alcohol in well-defined, transparent prisms, which become opaque a t 95--?@0", and melt to a black liquid a t 150-165"; if heated rapidly from 150°, no visible change occurs until about 160°, and then the substance begins to darken, and decomposes completely at 175--180".On treating a solution OF the substance in chloroform with a little bromine, the halogen is slowly absorbed, hydrogen bromide being evolved in small quantities; the product is a yellowish, crystalline sparingly soluble powder, but it was not analysed. When dibromhydrindone is treated with an alcoholic solution of sodium ethoxide instead of with alcoholic potash, other conditions re-REVIS AND KIPPING : DERIVATIVES OF a-HYDRINDONE.247 maining as already described, a similar blue or violef coloration is pro- duced, and a crystalline compound is easily isolated from the solution ; a sample, purified by recrystfillisation from alcohol and dried over sul- phuric acid, gave on analysis the following results. 0.1531 gave 0.3510 CO,, and 0.0570 H,O. C = 62.6 ; H = 4.1. C,,H,oBrO,*OEt requires C = 62.6; H = 3.9 per cent. From this it would seem that this product is closely related to indonylbromhydrindone, of which it is probably an ethoxy-derivative, but it was not examined very fully, as i t appeared to be of little interest. It crystallises from alcohol in colourless lamins, or in compact prisms which melt and decompose at 173-174"; it dissolves freely in hot chloroform, acetone, and benzene, but is only moderately soluble in ethylic acetate and alcohol, and comparat,ively sparingly soluble in cold ether.I~zdoizyl~~~yds.oxyhy~~~nlone, C:,sH,20,. When hydrindone is placed i n a stoppered bottle together with a considerable quantity of a warm solution of bromine in excess of caustic soda, it melts to a colourless oil, but on warming on the water bath f o r some hours, shaking from time to time, this oil gradually turns brown and becomes more viscous, and is finally partly converted into a light brown solid ; the latter is separated, washed with dilute alcohol to free it from all impurities, and recrystallised from boiling acetic acid ; for analysis, a sample was dried over sulphuric acid. 0.14'76 gave 0,4220 CO,, and 0,0644 H,O.0,1449 gave 0.4152 CO,, and 0.0602 H,O. C = 78.0; H = 4.8. C = 7S.2 ; H I= 4.6. CIsH,,O, requires C = 7S.9 ; H = 4.4 per cent. I n some experiments, apparently when the mixture was shaken vigorously for a short time, the oil was quickly converted into a white, flocculent substance, which was proved to be dibromhydrindone by its melting point and other properties, and also by analysis. Indonylhydroxyhydrindone is only sparingly soluble in boiling ethylic acetate, benzene, chloroform, and acetic acid, from all of which it is deposited in small, colourless needles on cooling ; it has not a definite melting point, but begins to turn yellow a t about 2303, and then gradually decomposes, although it does not liquefy even at 250". Attempts to throw some light on its constitution were unsuccessful, and the name assigned to it above is therefore merely a pro-visional one.Although it seems to interact with phenylhydrazine i n acetic acid solution, the product consisted of a tar from which a crystalline compound could not be isolated. It is only sparingly soluble in boil-248 REVIS AND KIPPING : DERIVATIVES OF WHYDRINDONE. ing acetic anhydride, from which it separates in part unchanged, but if the solution is boiled during some hours, crystals are not deposited on cooling, and a brown, resinous substance is formed. Action of Minesml Acids on Hgclrindojze-oxinie. The experiments which were made in order to compare the behnviour of hydrindone-oxime towards mineral acids with that of camphor- oxime, may be very briefly described, as they led to results of little interest.When hydrindone-oxime is heated with a little concentrated hydro- chloric acid, it slowly dissolves, but after some time crystals make their appearance, and the whole gradually becomes semi-solid. This crystalline product consists of a mixture of anhydrobishydrindone and truxene, and on allowing the hydrochloric acid filtrate t o stand, it deposits a small quantity of a colourless substance melting a t about 164' ; the latter consists of the hSdrochloride of hydrindone-oxime. When concentrated hydriodic acid is employed in the place of hydrochloric acid, the solution rapidly becomes dark brown, owing to the liberation of iodine, and a considerable quantity of anhydrobis- hydrindone is obtained after heating for a short time, but apparently truxene is not formed.These results shorn that hydrindone-oxime behaves quite differently from camphor-oxime, as indeed was to be expected, assuming that the formation of campholenic nitrile takes place in the manner suggested by Bredt. AZkcc Zi Derivatives of Iso~a~ti.osol~ycli.inclo?ze. In the course of some experiments on the reduction of isonitroao- hydrindone (Trans., l894,66,492),it was noticed that, on adding sodium amalgam to an alcoholic solution of the isonitroso-compound, a bright red or scarlet substance was gradually deposited as the pieces of amal- gam passed into solution; this substance proved t o be the sodium derivative of isonitrosohydrindone, but it may be more conveniently prepared simply by dissolving the isonitroso-compound in a little sodium ethoxide, and then precipitating with alcoholic ether ; the yellow, crystalline substance thus obtained was purified by recrysballisa- tion from moist ether, dried in the air, and analysed.0.1167 gave 0.0448 Na,SO,. Na = 12.5 per cent. C,H,NO,Na requires Na = 12.6 per cent. This sodium derivative crystallises from cold alcohol, and from moist ether in pale canary-yellow prisms, and it is moderately easily soluble in hot methylic alcohol, but insoluble, or nearly 80, in dry ether, chloro- form, and light petroleum. It dissolves freely in hot water, but it is at the same time decomposed to a greater or less extent into sodiumREVIS AND KIPPIKG : DERIVATIVES OF a-HYDRINDONE. 249 hydroxide and the isonitroso-compound ; on allowing the solution t o cool, it deposits crystals which are free from sodium, and which have the same melting point as isonitrosohydrindone.When the yellow, undried crystals of the sodium derivative are heated, they change, even at temperatures below loo", into a scarlet substance, and when previously dried they also undergo this transfor- mation, but, apparently, not at so low a temperature; when, on the other hand, the scarlet substance is kept at the ordinary temperature, it gradually assumes its original pale yellow colour. This change in colour is caused by a change in the crystalline form of the substance, and the following account includes the result of an examination of the crystallographic properties of the two modifications which was made for us by Mr.W. J. Pope. '6 The crystals consist of very small, thin, flattened, transparent orthorhombio needles, which shorn the forms (1001, ( O l O ) , and { O l l ) , and are elongated in the direction of the c-axis, and flattened in that of the a-axis ; they can, therefore, be examined through the faces of (100) only. They are pleochroic, appearing colourless in light the plane of polarisation of which is parallel t o the c-axis, and of a brilliant straw yellow colour in light polarised parallel to the b-axis. The optic axial plane is (OlO), and the a-axis is a positive bisectrix ; the optic axes lie outside the field of a &th immersion objective, and the interference figure is the normal orthorhombic one ; the angle (011) : (011) = l l O o , whence 6 : c = 1 : 0.7 (approx.)." '' On heating to 70 - SOo, the transparent, yellow needles are converted into a nearly opaque red or scarlet modification, although still preserv- ing their shape ; the new crystals show aggregate polarisation, and are unsuitable for further examination. By rapid heating, the yellow crystals may be obtained partially covered with red patches; in oil, this red modification is permanent for some days a t the ordinary temperature, but in the air it becomes wholly reconverted into the yellow one in the course of a few hours." It may be noted that the scarlet modification of the sodium deriva- tive may also be obtained directly from solution by boiling down an alcoholic solution of the compound until a considerable quantity of t h e substance has separated ; in this case, the temperature is above that at which the yellow modification is stable.An aqueous solution of the sodium derivative, to which a little sodium carbonate has been added, is of a pale yellow colour, but it becomes distinctly darker, and assumes a reddish-brown hue, on warming, becoming pale yellow again on cooling ; the cause of this phenomenon is, probably, dissociation. The potccssium derivative of isonitrosohydrindone closely resembles the sodium derivative in all ordinary properties, and, like the latter, exists in a yellow and in a bright red crystalline modification.250 REVIS AND KIPPING : DERIVATIVES OF U-HYDRINDONE. Hycli-indoneccxine. Hydrindone interacts readily with hydrazine, giving a beautiful, yellow, crystalline product, which is formed in accordance with the following equation : 2C,H,O + N,H, = C9H,:N*N:C,H, + 2H20.I n preparing this substance, an aqueous solution of excess of hydra- zine sulphate is added t o an alcoholic solution of the ketone, the mixture is warmed on the water bath, and small quantities of a solution of potassium hydroxide are added from time to time ; interaction soon occurs, and dirty yellow crystals are deposited. When no further separation takes place, the solution is filtered, and the residue washed with hot water, dried, and recrystallised from boiling benzene. The ketazine is thus obtained in lustrous needles or prisms having the colour of quinone ; a sample dried at loo3 gave the following result on analysis. 0-192s gave 0.5900 GO, and 0.1110 H20.C,,H,,N,requires C = 83.08 ; H = 6.15 per cent. The ketazine separates from chloroform in large, well-defined, flat prisms, and from methylic alcohol in compact rhomboidal plates, or in large, flat, fern-like forms ; it is very readily soluble in boiling chloro- form and acetic acid, and dissolves freely in hoiling acetone and benzene, but is only sparingly soluble in boiling methylic alcohol and in light petroleum. It melts at 164-165', when heated rather rapidly from looo, decomposing slightly, and turning a darker yellow. It seems t o be insoluble in cold hydrochloric acid, but on heating, it becomes lighter coloured, and then dissolves, being decomposed, apparently with formation of hgdrindone. C= 83.46 ; H= 6.39. A.naidohycl~iizdene ccizcl its Derivatives.Amidohydrindene (hydrindamine) has already been described by Konig (loc. cit.), who prepared it by reducing hydrindone-oxime with sodium amalgam in acetic acid solution; in making large quantities of this base, we have found that the following method leaves little to be desired. The oxime is dissolved in dilute acetic acid, a small quantity of sodium amalgam added, and the whole well shaken until the amalgam is decomposed, the solution being allowed to become warm by tho heat developed during the process; the addition of amalgam is then continued, the solution being kept acid, until no precipitate is produced on treating a portion with water. The filtered solution is then rendered strongly alkaline, the base distilled in a current of steam, collected in hydrochloric acid, and the solution of the hydro- chloride evaporated ; the salt is thus obtained in colourless crystals having the properties described by Konig (loc.c i t ) .REVIS AKD KIPPING : DERIVATIVES OF U-HYDRINDONE. 2-51 I1?,2idolzy/di.inde.e oxnlaie, 2C,H,,N,H2C20,, is obtained when the base is neutralised with a solution of oxalic acid, and the mixture evapo- rated ; i t crystallises from water in rosettes of opaque, white prisms, and is only moderately soluble in water and sparingly so in cold methylic alcohol. A sample was analysed with the following result. 0.2136 gave 0.5258 CO, and 0.1318 H,O. SC,H,,N,H,C,O, requires C: =-' 6'7.4 ; H = 6.7 per cent. Anzidohyclyindene acetate, prepared in a similar manner, is obtained as a syrup when the solution is evaporated, but it gradually solidifies to a mass of Iong, silky needles, and after having been spread on a porous plate, it may be recrystallised from cold water; it is thus obtained in beautiful, transparent prisms which seem to contain water of crystallisation, as they melt gradually from about 74" to loo", giving a colourless liquid which appears to boil a t about 140" ; if, after leaving the melted substlam% at this temperature for some time, i t is caused to crystallise by cooling, i t melts a t 113-115" when heated for the second time.Amidohyclriizclene zitra rate was obtained in the course of some experi- ments on the action of nitrous fumes on an ethereal solution of the base, and may also be obtained by treating the base with dilute nitric acid ; it crystsllises in long, colourless prisms and dissolves very readily in water, but is insoluble, or only sparingly soluble, in dry ether.BenxoyZc~mido~~ydi.in~Zene, C,H,*NH- CO*C,H,, is a well-defined, crys- tallised compound which serves as a means of identifying the base. It is very easily prepared by treating t h e base or its hydrochloride with 10 per cent. potassium hydroxide and benzoyl chloride in the usual manner ; the oily base is thus rapidly converted into a solid or pasty mass, which is separated by filtration, washed with dilute alcohol and recrystallised. C=67*5 ; H = 6 9 A sample was analysed with the following result. 0,2063 gave 0.6106 CO, and 0.1195 H,O. UIGH,,NO requires C = 81.0 ; H = 6.3 per cent. Benzoylamidohydrindene is readily soluble in alcohol, from which it crystallises in colourless, silky needles melting a t 142-143' ; it is very readily soluble in chloroform, ether, acetic acid, and ethylic acetate, but insoluble or nearly so in light petroleum.It dissolves in concentrated sulphuric acid, giving a dark red or crimson solution which has a green fluorescence, but on adding water the colour vanishes. Benx~~ide?aeamidohycl~.i?~zdeize, C,H,N :CH* C,H,, is formed when the base is treated with an alcoholic solution of benzaldehyde and a few drops of potash; a better method of preparation, perhaps, is to dissolve the base, together with a slight excess of benzaldehyde, in C = 80.7 ; H = 6.4.252 REVIS AND IilPPISG : DERIVATIVES O F a-HYDRINDOSE. ether, keeping the solution at the ordinary temperature for about 6 hours. The ether is then evaporated, the residue boiled with water until free from benzaldehyde, and then recrystallised from dilute alcohol; the compound is thus obtained in clusters of transparent prisms which melt a t 74-75". 0.2506 gave 05'914 CO, and 0,1534 H,O. C = 86.2 ; H = 6 8. A great many experiments were made with the object of converting amidohydrindene into the corresponding hydroxy-compound by the action of nitrous acid, but although the conditions were modified in various ways, it was found impossible to avoid the production of a large quantity of resinous matter, and, as KGnig had previously stated,?the yield of hydroxy-compound was invariably very poor. Equally~pooresults were obtained on passing the fumes evolved by the action of nitric acid on arsenious oxide into an ethereal solution of the base ; the nitmte, already described, separated from the ethereal solution in brownish crystals, but otherwise little action seemed to occur. The direct conversion of hydrindone into the alcohol also seems to be impracticable. Konig, who tried the action of varions reducing agents, obtained in all cases a considerable quantity of a crystalline pinacone, and our experience has been the same. Even a neutral reducing agent, such as aluminium amalgam, fails to bring about the desired result, as in this case also the principal product is a crystalline compound melting a t about 143", evidently the pinacone. It may be mentioned also that this pinacone is formed by the action of nitrous acid on the base, a somewhat unusual reaction, which is probably the I esult of oxidation. We can also confirm KGnig's statement that the (impure) alcohol, prepared by treating the base with nitrous acid and purified by distillation in a current of steam, is very easily converted into resinous compounds when it is warmed with hydriodic acid. This being the case, it seemed useless to continue these attempts t o conrei t the ketone into hydrindene. C,,H,,N requires C = 86.9, 73 = 6.8 per cent. CHEMICAL I)EPdRTMENT, CENTRAL TECHRICAL COLLEGE, CITY ASD GUILDS OF LOXDON ISSI'ITUTE.