KENNER : THE HEDUCTlON PRODUCTS, ETC. 2685CCLI.-- The Reduction Products of Ethyl Hydrindenc-2 2-dicarboxylate.By JAMES KENNER.IN a previous communication it was suggested that the ease withwhich cyclic condensation occurs should be modified by the presencein a chain of certain carbon atoms which, being already membersof a closed ring, have the directions of their valencies to'someextent determined (Kenner and Turner, T., 1911, 99, 2102). Theinvestigation now to be described was undertaken with the objectof studying ring-formation from compounds, the molecules ofwhich contain one carbon atom fulfilling this condition; in otherwords, the preparation of sp-ro-compounds was to be attempted.Since substituent groups are also known to be important factorsin determining the facility of formation, and stability, of cycliccompounds it appeared that, among spire-compounds, .the hydro-carbons would furnish the most decisive evidence of the influencesreferred to in the preceding paragraph.These and other considera-tions suggested the application t o ethyl hydrindene-2 : 2-dicarboxy-late of Bouveault and Blanc's method of reduction by means of* These rotations are somewhat lower than those previously published. It hasnot been thought necessary to investigate the cause of the discrepancy, as this paperdeals mainly with optical dispersive power, which is only very slightly affected bythe smsll discrepancy2686 KENNER : THE REDUCTION PRODUCTS OFsodium and ethyl alcohol, in the hope of preparing oo’-dzhydrozy-2 : 2-dimethylhydrindene (I), from which the hydrocarbon (11)C6H,<CHqC<FH, CH, CH,’w.1might subsequently be obtained.It had previously been shown by Bouveault and Blanc that ethyldiisobutylmalonate, whilst furnishing a certain amount of theexpected glycol, was t o a considerable extent decomposed in thefollowing way under the influence of sodium ethoxide formedduring the reaction :the ethyl isovalerate being then reduced in the normal manner(Bull.SCC, chim., 1904, [iii], 3 1, 1203). Ethyl hydrindenedicarb-oxylate had, however, been shown by Thole and Thorpe to bequite stable towards sodium ethoxide a t the ordinary temperature(T., 1911, 99, 2186), and the hope was therefore entertained thateven a t the higher temperature to be used in these experiments thetype of decomposition observed by Bouveault and Blanc might notassert itself in a marked degree.This expectation, however, wasnot realised, for the yield of the glycol (I) was disappointinglysmall, being less than 3 per cent. of the calculated. More than40 per cent. of tho ethyl hydrindenedicarboxylate was convertedinto 2-htydroxymethylhydr~n~ene (111), the remainder being re-covered in the form of a mixture of hydrindene-mono- and di-carb-oxylic acids, in which the former largely predominated :(C4H,),C(COzEt)z + CZH,*OH ++ (C4H,),CH*COzEt + CO(OEt),,Initially, therefore, the dicarboxylic ester was almost entirelyconverted into the monocarboxylic ester, and, in the author’sopinion, this reaction must be ascribed to spatial causes, whichwill bO discussed lahr.It is probable that such influences also playa part, although possibly a subordinate one, in promoting thedecompositions discussed by Thole and Thorpe (loc. cit .).2-Hydroxymethylhydrindene was readily converted by the usualmeans into 2-bromonzethylhydrindei~e (IV), the reactions of whicETHYL HYDRINDENE-2 8-DICARBOXYLATE. 2687invited investigation, because it has frequently been observed thatthe bromine atom in derivatives of this type is remarkably inert.Thus Perkin and Pope found that 1-methyl-4-bromomethylcyclo-hexane (V) was converbd into the cyanide only with considerabledifficulty (T., 1908, 93, 1079). SimiIar relationships were dis-covered in tha present instance. The bromo-compound was un-changed after prolonged boiling with amalgamated zinc and hydro-chloric acid, in spite of the efficiency of this reducing agent (Clem-mensen, Ber., 1913, 46, 1837; 1914, 47, 51, 681).Interaction ofthe bromo-compound and ethyl sodiomalonate in alcoholic solutionafter ten hours a t the boiling point resulted in the production ofonly about 65 per cent. of tlie calculated amount of ethyl 2-hydrin-dylrnethylmnlonate (VI) :*C 6 H 4 < ~ ~ 2 > C H - C H2-CH (CO,Et\,The formation of 2-phthalimi?zomethylhydrindene (VII) by heat-ing the bromo-derivative with potassium phthalimide at 180-200°for nine holm was similarly incomplete:2(VI.)C6H4<g: 2>CH*CH 2*N<Eg>C,H4(VII.)The contrast between the inertia of the bromine atom in suchcompounds and its activity in, f o r inst'ance, benzyl bromide, isworthy of some comment, and is obviously in some way connectedwith t,he difference between the saturated and the unsaturatedconditions of the cyclic structures present in the two types ofcompounds.If, however, benzyl bromide be represented by theformula (VIII), in Flurscheim's notation, it would appear to followas a striking consequence that Perkin and Pope's 1-methyl-4-bromo-methylcyclohexane is t o be represented by the formula I X :(VIII.) (IX).The similar inertia of the bromine atoms in tetrabromotetra-methylmethane (Perkin and Simonsen, T., 1905, 87, 161; Fecht,Ber., 1907, 40, 3884) would then find expression in the formula X :Br-CH CH --BrBr--CH;>C<C +I3 L'(X-2688 KENNER: THE REDUCTION PRODUCTS OFThe facts symbolised by these formuh are illustrative of theinfluences referred to a t t’he commencement of this paper, and, inthe author’s opinion, they are all explicable by a thorough appli-cation of Baeyer’s strain theory.For it is a t once clear that thenormal relative positions of substituents known to exert sterichindrance, such as methyl (or bromomethyl) and carboxyl groups,may, when they are attached to the same carbon atom, be compar-able, in regard to this atom, with those of the carbon atoms in,for instance, a cyclohexane or a cycloheptane ring. Then, adoptingWerne’r’s conception of the uniform spherical distribution of affinityround a carbon atom (“Beitrage zur Theorie der Affinitiit undValenz,” Zurich, 189l), we see that, if aal in XI represent insection the zones of affinity appropriated by two univalent group-ings in the plane of the paper when “the angle between theirvalencies is 109O281,” an increase in this angle will cause an altera-tion in the relative position of the zones, which will now be repre-sented by XII:(XI.) (XU.)I n this manner a certain amount of affinity, corresponding withthe region (b), will be left unsatisfied, and the extent of this regionis a measure of the “strain,” in Baeyer’s terminology.I f the othergroups attached to the carbon atoms be free t o move, they willprobably so adjust themselves as partly to engage the valency thusleft free because the change in position of the zones aa’ involvesan incursion into the zones of affinity previously available for them.I n the following paragraphs, the attempt is made to apply theseconsiderat,ions to the cases in which ( a ) two of the groups attachedto a carbon atom are components of the same) cyclic system, whilstthe other two are groups of large molecular volume.I n this case,the motion of the former groups is restricted, and the affinity repre-sented by b will then remain free and available t o a greater or lessextent as partial valency to an atom situated above or below theplane of the paper and, for example, coplanar with the groupsappropriating a d .Thus, in the case of ethyl hydrindenedicarboxylate there will beresidual affinity on the quaternary carbon atom, and, owing to theencroachment of the carbethoxy-groups on the zones of affinitETHYL HYDRINDENE-8 8-DICARBOXYLATE.2689normally available for the two carbon atoms of the hydrindene ring,one or more of these groups will obtain less than its proper shareof affinity. This deduction is in agreement with the experimentalevidence just advanced, according to which we may conclude thatthe ester is more adequately represented by the formula XIII:(XIV.)Similar considerations probably supply an explanation of anumber of reactions met with in the chemistry of cyclic compounds.I n illustration of this may be cited the change of carone intocarvenone by distillation (Baeyer, Ber., 1894, 27, 1917), and intobromo- and hydroxy-menthanone by absorption of the elementsof hydrogen bromide or water (ibid., p.1920); the isomerisationof carylamine hydrochloride into vestrylamine hydrochloride(Baeyer, Ber., 1894, 27, 3486); the disruption of the bridgein dimethyldicy clopentanonecarboxylic acid by reduction (Perkin,Thorpe, and Walker, T., 1901, 79, 729); the addition of theelements of hydrogen bromide t o a-camphylic acid (Perkin, T.,1903, 83, 842); and the various reactions by which the bridge inthe camphor molecule is broken between two quaternary carbonatoms (see Aschan, “ Konstitution des Kamphers,” Braunschweig,1903, p. 79).The reactions of certain other compounds are illustrative ofanother mode of relieving the stlress on the quaternary carbon atom,namely, the replacement of two single bonds by a double bond.This results in a smaller demand being made on the affinity of thecentral carbon atom.Thus Wallach has shown that ethyl cyclo-hexan-1-01-1-acetate on hydrolysis is partly converted into cyclo-hexanone, accompanied by some’ cyclohexanol (a hydrogen atomhaving displaced a group of large molecular volume). Further,dehydration of the ester or of the acid is easily carried out, andresults in the formation of Al-cyclohexeneacetic acid or of carboxy-methylenwyclohexane, according t o the agent employed. Indeed,the initial condensation product of 1 : 5-dimethyl-Al-cycZohexen-3-one cannot be isolated, but passes over at onoe into 1 : 5-dimethyl-A1:3-cyclohexadienyl-3-acetic acid (Annalen, 1900, 3 14, 147 ; 1902,323, 135; 1905, 343, 40, 347, 316; 1908, 360, 26).That thesereactions are not due t o the presence of the hydroxyl group assuch is shown by a remarkable instance of an analogous kind, com-municated to the author by Prof. J. F. Thorpe. Ethyl cyclo-hexane-4-dibromodiacetate (XIV) when boiled with dilute potass-ium hydroxide solution is converted into carboxymethylenecyclo2690 KENNER: THE REDUCTION PR0I)UcTS OFhexane (XV), although whe,n it is dropped into concentratedaqueous potassium hydroxide a t 130° hhe acid (XVI) is produced :It is obvious that similar conditions will prevail when, as in theinstances quoted above, three o r four separate groups of largemolecular volume are attached t o a single carbon atom.When, as in the molecule of cyclopropane-1 : l-dicarboxylic acid,the cyclic structure is such that the '' angle between two valencies "of the quaternary carbon atom is less than 109O28/, the effectsjust discussed will be intensified.Hence this acid and 1 :l-di-methylcyclopropane are almost comparable with unsaturatedcompounds in the readiness with which they take part in addi-tive reactions, and the general conclusions of Kotz ( J . p. Chem.,1903, [ii], 68, 174) in regard to the derivatives of cyclopropaneare in agreement with the statement just made. Further,Radulescu's observation that the acid (XVII) is stable towardshalogen hydrides (Be?., 1909, 42, 2771; 1911, 44, 1018)appears to be direct evidence in favour of the suggestion that itscarbonyl groups are differently situated with regard t o the centralcarbon atom from those in cyclopropane-1 : l-dicarboxylic acid :7H2 CC*YH=CO,H(XYII.)The rearrangement of derivatives of ethylene oxide into those ofacetaldehyde are instances of a similar nature among heterocycliccompounds (Fourneau and Tiffeneau, Compt.rend., 1905, 141,662; Klages, Ber., 1905, 38, 1969; Klages and Kessler, Ber., 1906,39, 1753):>c<co-cH.co,H CH2Both pairs of valencies attached t40 the carbon atom ( b ) areinclined to one another a t angles less than 109O28/, when similarinstability of the molecule may be expected. Thus the followingtable shows in tho case of the central carbon atom of the spiro-compound, (CH2)J2(CH2),, the angle between a valency of the(x + 1)-membered ring and one of the (y + 1)-membered ring :Y+1x + l 2 3 42 180' 150' 136"3 139 12ETHYL HYDRINDENE-2 : 2-DICARBOXYLATE.2691The magnitude of these angles indicates that, considerableamounts of unsatdsfied affinity will exist between the zones corre-sponding with the valencies in question. Consequently, compoundsof this type may be very difficult to isolate, and, when obtained,very liable to undergo change. Thus Dimroth and Feuchter wereunable to prepare an allene derivative from the compounds XVIIIand XIX (BeT., 1903, 36, 2238; compare Ipatiev, J . pr. Chem.,1899, [ii], 59, 517):C02Et cGH6>C: CCl*CH,* CH,(XVIII.)CC), CeH5>CH*CCI Et :CH*CH,(XIX.)Similarly, ethyl allenetet,racarboxylate (XX), which is onlyobtained by heating the initial product (XXI) of the action ofethyl sodiomalonate on carbon tetrachloride, absorbs two molecularproportions of water when exposed in a moist atmosphere (Zelinskiand Doroschevski, Ber., 1894, 27, 3376):(XX.j.-The action of alcoholic potassium hydroxide on iodomethylcyclo-propane leads to the production of erythrene, presumably owing tothe rearrangement of methylenecyclopropane (Dem janov, J . Buss.Phys. Che.m. SOC., 1903, 35, 375):Also, Favorski and Batalin have recently shbwn (Ber., 1914,47, 1648) that Gustavson was mistaken in attributing the consti-tution of an ethylidenecyclopropane to a compound he had preparedin an analogous manner (Compt. rend., 1896, 123, 242). Further,the sole product of dehydration of cyclopropyldimethylcarbinol isP-cyclopropylisopropylene (XXII), notwithstanding the fact thatOH.C(CB,)2*CH<yH2 + CH2:C(CH,)-CH<FH2C*2 CH2(xxIr.1dimethylisopropylcarbinol furnished the isomeric olefines (XXIIIand XXIV) in the proportion of three to one (Henry, Compt.rend.,1908, 147, 557) :OH*C(CH,),*CH(CH,), --+ C(CH,),:C(CH,), and(XXIII.)CH,:C(CH,)*CH(CH,),.(XXI v. 2692 KENNER: THE REDUCTION PRODUCTS OFThe production of methylenecycZobut*ane (XXV), in place ofspiropentane (XXVI) , from t2etrabromotetramethylmethane isdoubtless t o be ascribed to similar causes (Demjanov, Ber., 1908,41, 915; Favorski and Batalin, Zoc. cit.; compare Gustavsonand Bulatov, J . pr. Chem., 1896, [ii], 54, 97; 56, 93; Fecht,Zoc. c i t . ; Zelinski, Ber., 1913, 46, 170):(XXV.) (XXVI.)Neither this hydrocarbon nor cyclobutanone (Kishner, J . Buss.Phys. Chem. SOC., 1905, 37, 106; 1907, 39, 922) exhibits anytendency towards the breaking down of the f our-membered ring,but it is significant that cyclobutane-1 : 3-dione behaves as thoughit were represented by the formula XXPII (Chick and Wilsmore,T., 1910, 97, 1982):CH,*C<'I O:C*CH,----.(XXVII.)The illustrations thus brought forward are not intended to beexhaustive, but suffice to indicate the aspect from which, in theauthor's opinion, the study of spko-compounds should beapproached. The quaternary carbon atom is not per se a sourceof weakness, this being conditioned by the distortion of its valenciesfrom their normal positions.Finally, it may be mentioned that experiments have also beeninitiated with a view, on the one hand, to the synthesis of thecompound (XXVIII) by the condensation of the chloride of hydrin-den62 :%dicarboxylic acid with benzene, and, on the other, t o the(X XVIII.)preparation of reduction products of ethyl cyclohexanediacetate,' from which sp*ro-compounds might be prepared. The investigationin this direction has, however, only just been commenced, and thememure of success attained is indicated in the experimental portionof this paper.EXPERIMENTAL.Reduction of Ethyl Hydrindene-2 : 2-dicarboxylate.Sodium (30 grams), cut into pieces the size of a pea, was placedin a large flask, fitted with a long, upright condenser and a tap-funnel, and the flask was heated to 80° in an oil-bath.A solutioETHYL HY DRINDENE-8 : 2-DICARBOXYLA1 E. 2693in absolute alcohol of ethyl hydrindenedicarboxylate (23 grams),previously purified by distillation under diminished pressure(100 c.c.), was then run from the tap-funnel on to the sodium asrapidly as possible, consistent with efficient action of the condenser.Th0 temperature of the oil-bath was then raised t o 130°, and afurther quantity of alcohol (100 c.c.) gradually added in the courseof two hours. A t the end of five hours from the experiment anyunreduced ester was hydrolysed by the gradual addition of watert o the mixture. The product was then cooled, considerably diluted,and treated with sufficient sulphuric acid to leave the solutionweakly alkaline.By exhaustive extraction with ether the mixtureof reduction products was removed, whilst hydrindenemonocarb-oxylic acid (5.5 grams) could be recovered by subsequent acidificn-tion of the aqueous solution.The ethereal extracts, after treatment in the usual manner,furnished an oil, which was distilled under diminished pressure.I n this manner a large fraction (5.5 grams) was obtained, whichboiled a t about 140"/11 mm. and solidified a t the ordinary tem-perature.2-Hydroxymethyll~ydrindene (111), obtained in this way, has acharacteristic agreeable odour, and consists of prismatic crystals,which melt a t 3 3 O and boil a t 139-140°/11 mm. It is readilysoluble in niost organic solvents, but only sparingly so in lightpetroleum (b. p. 40-50°), and may be crystallised from thissolvent :0.1490 gave 0.4424 CO, and 0.1078 H,O.C = 80.98; H = 8.04.C,,Hl20 requires C = 81.07 ; H = 8.11 per cent.The phenylurethane, C,H,<CB2>CH*CH2*O*C0.NH.C,W,, CH was2prepared by heating a solution of molecular proportions of thecarbinol and phenylcarbimide in light petroleum (b. p. 90-110O).After crystallisation, it melted a t 99.5O :0.2188 gave 10.2 C.C. Nz a t 1l0 and 755 mm. N=5*59.CI7Hl7O2N requires N = 5-24 per cent.w wf-Dihydroxy-2 : 2-dimet~~lhtyd~~nde?ze (I) was obtained by dis-tilling the united residues from four of the above preparations of2-hydroxymethylhydrindene. A colourless oil passed over a t about200°/15 mm., and rapidly solidified. On the addition of lightpetroleum (b. p.90-110O) to its solution in ethyl alcohol, small,hexagonal prisms, melting a t 112.5O, separated. The yield was1.5 grams:0.1568 gave 0.4266 CO, and 0.1090 H20. C =74*20; H =7*73.C,,H,,O, requires C = 74.16 ; H = 7.86 per cent.VOL. cv. 8 8694 KENNER: THE REDUCTlON PRODUCTS OFR eductio 1% of Ethyl Hydrindene-2-car b oxylate.A solution of tha ester (20 grams) in alcohol (100 c.c.) was addedto sodium (18 grams> in precisely the same manner as alreadydescribed for fhe previous case, alcohol (20 c.c.) being subsequentlyadded. The yield of carbinol was 8.5 grams.2-Hydroxymethylhydrindene (10 grams), having been added toa solution of chromic acid (4.2 grams) in 10 per cent. sulphuricacid (75 grams), thp, mixture was heat'ed on the water-bath f o r twohours.The ?thereal extract of the cooled solution was washed withsodiuni carbonate solution, and then shaken with concentratedsodium hydrogen sulphite solution. The aldehyde, isolated in pooryield frcm this solution in the usual manner by decompositionwith sodium hydrogen carbonate, was a fairly mobile oil, boilinga t 122O/12 mm., which did not, solidify. It readily underwentoxidation on exposure, and its odour also characterised it as analiphatic aldehyde :0.1440 gave 0.4325 CO, and 0.0878 H,O. C=81*91; H=6*77.CloHloO requires C = 82.19 ; H = 6.85 per cent.The semicarbasone, prepared in the usual manner, readily dis-solved in alcohol, and separated from this solvent in radiate masseaof small needles melting at 174O:0.1120 gave 20.6 C.C.N, a t 17O and 730 mm. N=20*90.C,,H,,ON, requires N = 20.69 per cent.2-Bronzom e t h y l h ydrindene (IV) .This compound was easily prepared by heating a solution of2-hydroxymethylhydrindene (35 grams) in glacial acetic acid,saturated a t Oo with hydrogen bromide (50 c.c.), a t 100-1200 forthree and a-half hours.The compoulnd boiled a t 132O/11 mm., and solidified a t lowtemperatures t o rnassa of magnificent prisms, melting a t 210. Itsodour was characteristic and reminiscent of aniseeld :0.1742 gave 0.3646 CO, and 0'0812 H,O. C=57-08; H=5*19.It was recovered unchanged after being boiled for ten hours withC,,H,,Br requires C = 56.87 ; H = 5.21 per cent.amalgamated zinc and dilute hydrochloric acidETHYL HYDRINDENE-2 2-DICARBOXY LATE.2695Coitdensatiorh of 2-Bromomethylhydrindene with Ethyl Mrilonate.Ethyl malonab (6.4 grams) and the bromo-compound (8.4 grams)were successively added to a solution of sodium (0.9 gram) in alcohol(14 c.c.). A t the temperature of the water-bath a separation ofsodium bromide soon commenced, and after ten hours the productwas worked up in the usual manner By distillation underdiminished pressure, well-defined fractions of ethyl malonate,bromomethylhydrindene, and finally of the desired ester (7.5 grams)were obtained.Ethyl 2-hydrindylmethylmalonate (VI) is a colourless liquid,which boils a t 211°/15 mm., and does not solidify even when cooledin a freezing mixture:0.1594 gave 0.4092 CO, and 0.1074 H,O. C = 70.00; H = 7.48.C1,H2,04 requires C = 70.35 ; H= 7-59 per cent.The corresponding acid was prepared by hydrolysis with alcoholicpotassium hydroxide, and separated from its solution in alcohol inclusters of small, transparent plates, melting a t 174O :0.1746 gave 0.4285 CO, and 0.0965 H20.C = 66.93 ; H = 6.14.C13H1404 requires C = 66.66 ; H = 6.00 per cent.Its barium, calcium, lead, tin, and ferric salts are insoluble inh o t water, whilst its mag,zesiurn, copper, and cobalt salts are solublein cold water.ThO dihydrazide, C,H4<~~2>CH*CH,*CH(CO*NH*NH2)z, crys-0.1868 gave 35.6 C.C. N, at 2 3 O and 747 mm. N=21*6.2tallises from alcoholic solution in silky needles melting a t 177O :C13Hl,0,N4 requires N = 21.4 per cent.8-2-Hydrindylpropionic acid, C,H4<E::>CH=CH2*CHz*C02H,was prepared by heating the a'bovel acid a t 190° until the evolutionof carbon dioxide had ceased.It was readily soluble in benzene,but sparingly so in hot light petroleum (b. p. 90-llOo), andseparated from a mixture of these solvents in small plates meltinga t 120O:0.1758 gave' 0.4902 CO, and 0.1166 H20. C=76.05; H=7*37.0.2474 required 14.6 C.C. N / 10-NaOH. Equivalent = 189.7.C12H14O2 requires C = 75-79 ; H = 7-37 per cent. M.W. = 190.Its barium and magnesium salts are soluble in cold water, whilstits calcium salt is sparingly soluble, and separates from its solutionin hot water in needles. Its ferric, copper, and cobalt salts areinsoluble in hot water, its lead and tin salts sparingly so, and itsmercuric salt turns yellow when boiled with water2696 KENNER : THE REDUCTION PRODUCTS OF%Phthalimii? omethtyEltydriiideit e (VII) .An intimate mixture of 2-bromomethylhydrindene (10 grams)with potassium phthalimids (9 grams) was heated a t 180-200° fornine hours in an apparatus provided with a reflux tube.Themixtnre solidified on cooling, and required to be finely powderedbefore adherent oily matter could be removed by repeated diges-tion with hot light ptroleum (b. p. 90-110O). Potassium bromidehaving then been removed by extraction with hot water, the residuewas crystallised from glacis1 acetic acid. The compound separatedin small, slender prisms, which were usually somewhat discoloured,and melted a t 174O. The yield was 57 per cent. of the theoretical,and was not improved by carrying out the condensation in thepresence of sodium iodide:0.1984 gave 9.0 C.C.N, a t 16O and 752 mm. N=5*30.C18H1502N requires N = 5-05 per cent#.2-EEydrkdylrn e t h y la mk,e, C6H4<g2> CH*CH2*NH2.Phthaliminomethylhydrindene (8 grams) was heated with con-centrated hydrochloric acid" (35 c.c.) a t 180-200° for six hours,and the product was then heated in the usual manner. A largeproportion of the hthalimino-derivative remained unchanged, butfour such experiments furnished a sufficient quantity of the base,boiling a t 248O, t o perlr.it of its characterisation.The hydrochloride separated from its solution in dilute hydro-chloric acid & thin plates with a satiny-lustre, melting and decom-posing a t 258-260O:0.1890 gave 12.4 C.C.N, a t 15O and 751 mm.The platinichloride was obtained as a yellow powder, whichN=7*68.CloH13N,HCl requires N = 7.63 per cent.decomposed a t 233O:0.3614 gave 0.1008 Pt. Pt=27*89.(C,,H13N)2,H2PtC16 requires Pt = 27.70 per cent.The iodide, sulphate, oxalate, and phosphate are readily solublein water, whilst the carbonate (prismatic needles) and thedichromate (orange, prismatic needles) are soluble in hot water.2-Phenylthiocarbamidome t h ylhydrilbdene,c , H , < ~ ~ CH ;>CH-CH,-N H~S.NH*C,H,,crystallises from alcohol in hexagonal plates melting a t 145O:0.1760 gave 15.5 C.C. N, a t 16O and 745 mm. N=10-20.CI7Hl8N2S requires N = 9.93 per centETHYL HYDRINDENE-8 : 8-DICARBOXYLATE. 2697The Chloride of Bydrindene-2 : 2-dicarboxylic Acid,C,H,<gg2>C( COCI),.2This compound was prepared by the interaction of the calculatedamounts of hydrindenedicarboxylic acid and phosphorus penta-chloride. It boiled at 173-175°/20 mm., and solidified a t theordinary temperature. It crystallised from light petroleum (b. p.60--80°) in clusters of reckangular plates, which melted a t 45O,and did not exhibit any marked tendency towards decompositionin contact with the atlmosphere ;0.2516 gave 0,2952 AgC1. C1= 29.02.C,,H,O2C1, requires C1= 29.22 per cent..An attempt was made to condense this compound with benzeneunder the conditions employed by Freund (Annalen, 1910, 373,310) in the case of diethylmalonyl chloride. It was found that, asin the latter case, the liquor obtained by steam distillation of theproduct was coloured green, and a small quantity of golden-yellowcrystals was obt'ained by extraction with ether. There can, there-fore, be no doubt that the reaction took the desired course.Reduction, of Ethyl cycloHexnnediacetate.This operation was carried out in the manner already describedin the case of ethyl hpdrindenedjcarboxylate. The oil obtainedboiled indefinitely, but small quantities of solid matter separatedfrom the later fractions, boiiing a t 195-200°/21 mm. This productwas sparingly soluble in light petroleum (b. p. 60-80°), andmoderately so in benzene. By crystallisation from this solventlezflets, melting at 123O, were obtained:0.1106 gave 0.2848 CO, and 0.1046 H,O. C = 70.23 ; H = 10.51.C,,H,,O, requires C = 70.06 ; H = 10.06 per cent.The author hopes to prosecute his investigations in the directioxindicated a9 soon as circumstances permit a resumption of theexperiments.THE UNIVERSITY,f!i HEFFTELTI