ORGANIC CHEMISTRY.PART I.-ALIPHATIC DIVISION.THE fact that a- considerable proportion of the original literaturewas no longer available during the second half of the past yearhas had comparatively little effect on the compilation of thissection of the Annual Report. The tot,al number of papers re-viewed has been about the average, and the quality of the workdescribed has been well maintained, but, for obvious reasons, verylittle progress has been made in some of the new developmentswhich were described last year. Nevertheless, many importantadvances have been made on systematic lines, although, in somebranches of the subject, a number of novel results have beendescribed, which are isolated observations, and therefore unsuit-able for general discussion. Taking these factors into account,it has been considered advisable t o limit the number of topicsreviewed in the Report, and t o deal with these in somewhatgreater detail than is possible in normal circumstances.Optical Activity.The preparation of optically active compounds and the continua-tion of attempts to correlate rotatory power with constitution arestill being continued with unabated vigour, so that new facts andnew theories have rapidly accumulated during the past year.Asquantitative work on optical activity is naturally carried out oncompounds of simple type, the subject is one which may be appro-priately discussed under the general heading of aliphatic com-pounds. Considering the number of papers to be reviewed, atten-tion will be restricted in the following account to cases in whichactivity is dependent on the presence of ‘ I asymmetric carbonatoms ” only.I n the first place, attention may be directed to the satisfactorysolution of a problem of long standing.The question has oftenbeen raised as to the simplest possible type of organic compoundwhich should be capabIe of displaying optical activity, and664 ANKUAL REPORTS ON THE PROGRESS OF CHEMISTRY.it has now been shown that when a single carbon atom is unitedto three different elementary atoms and to an inorganic radicle, theresulting molecule can exist in active forms. The compound tofurnish this interesting result was chloroiodomethanesulphonic ... acid,c1I>C<&,H~and it is noteworthy that the activity, which was measured on theammonium salt, was not impaired by the action of the usualracemising agents.*Turning to another problem of fundamental importance, a mostinstructive transformation is reported by Fischer, who has, afterconsiderable difficulty, succeeded in interchanging two groupsattached to the same asymmetric carbon atom, so that an opticalinversion has now been established in which all the intermediatecompounds have been isolated.* d-isoPropylmalonamic acid was,in the first place, converted into the methyl ester, and the amino-group expelled by the action of nitrous acid.Thereafter, throughthe intermediate formation of the corresponding hydrazidic acid,the esteric methoxyl group was replaced by the amino-group, and,in this way, the desired interchange of two groups was completedwith the result, already indicated, that the Z-isomeride of theoriginal acid was formed.Inspection of the structural schemeillustrating this cycle of changes will show that the attachment ofthe four groups to the asymmetric carbon atom remains undisturbedduring the various reactions, and that a t no stage is any group,directly attached to the asymmetric atom, completely removed :H CO,H H CO,EI7 --C,H,>U<CO,MC? --+ C,H,>C<C'O.NH2-- -+ B.The change from A to B thus constitutes an inversion, as indi-cated by the looped arrow, a convenient expression which will befound useful in blackboard illustration.3Reference may, a t this stage, be made to another idea whichmay likewise be of service in teaching, and that is a revival of thesuggestion that racemisation, even in the case of acids, is merely aparticular case of tautomeric change, and is due to the oscillationof a labile hydrogen atom.* There is much to be said for thisM'.J. Pope and J. Read, T., 1914, 105, 811.E. Fischer and F. Brauns, Xitzungsber. K. Akad. Wiss. Berlin, 1914, 714 ;0. Rothe, Ber., 1914, 47, 843 ; A . , i, 538.A , , i, 942. * P. F. Frankland, T., 1913, 103, 741ORGANIC CHEMIS'I'KY. 65suggestion , particularly as the best defined examples of auto-racemisation occur in the case of compounds which are definitelytautonieric, but nevertheless an explanation of racemisation whichwould be perfectly general in its application is still t o be found.Several publications of the past year deal with the allocation ofdefinite configurations t o a number of aliphatic compounds.It is,of course, notorious that the vagaries of the Walden inversion muststill render uncertain speculations on configuration which are basedupon reactions involving substitution and replacement of groups.At the same time, an interesting series of transformations has re-sulted in the determination, with some degree of certainty, of theconfigurations to be assigned t o the active glyceric and lactic acids.5Starting with the amide of I-malic acid, t'he corresponding ainicacid was prepared and converted into I-isoserine,C02H*CH(OH)*CH,*NH,.From this, in turn, d-lactic and d-glyceric acids were obtained, theformer by a somewhat indirect process, and the claim is made thatthe respective configurations of the various compounds involvedmust be as represented below:p , HOH*F*Hyo2H 7*2H /+ CH,*OHd-Glyceric acid.CH2*C02H CH2*NH2OH*$I*H --+ OH*$!*Hyo,* Z-Malic acid.I-isoserine. 1 90H.y.HCJ%d-Lactic acid.These views can certainly be supported by a number of argu-ments, and the scheme of the research is ingenious, but, consideringthe nature and variety of the reagents involved, i t seems unlikelythat the transformations are entirely unaccompanied by changes inconfiguration.I n reviewing the numerous publications which have appearedduring the past twenty years regarding the relationship betweenconstitution and molecular rotation, one cannot fail to be impressedwith the great changes which have taken place in the experimentalmethods employed arid in the selection of suitable compounds forexamination.As a result of many patient investigations, duerecognition is now given in accurate work t o the effect of tempera-ture, concentration, the nature of the solvents used, dispersioneffects, and other important factors, whilst the attention now paidt o racemisation affords a better guarantee that optically pure com-K. Freudenberg, Ber., 1914, 47, 2027 ; A . , i, 924.REP.-VOL, XI. I66 ANNUAL REPORTS ON THE PROGRESS OF CHEMISTRY.pounds are used. Perhaps the most prominent feature of recentwork in this field is the large number of optically active compoundswhich have been examined, and the completeness with which homo-logous series of these cornpounds have been built up on systematiclines.The pages of the Journal bear striking testimony to thispoint in the consecutive investigations described by Pickard andEenyon. Obviously, work of this magnitude and detail can onlybe studied and appreciated by reference to the original papers, buta brief summary of the present position of the subject may begiven, as a number of definite advances have recently been made.I n the first place, i t should be mentioned that, in continuation ofhis previous work, Lowry has examined the magnetic rotatory dis-persion of a large number of organic compounds which, for themost part, are of simple type.6 It is found that, generally speak-ing, the values determined are remarkably constant in homologousseries.On the other hand, optical rotatory dispersion is morevariable,’ although, in the case of disubstituted carbinols of thealiphatic serim, the value becomes constant when, in a homologousseries, the fraction of the molecule in which the chain is lengthenedbecomes the heaviest part of the asymmetric system. Further, ithas been shown that the formula ct=k/(h2-A$), which can beapplied generally t o express the rotatory dispersion of organic com-pounds in the homogeneous state, is consistent with the valuesfound for active carbinols a t moderate temperatures, and in somecases a t all temperatures up to the boiling point of the liquids.8The fact that the rotatory dispersions of esters d o not conformuniformly to the above formula must be accepted as indicatingthat ethereal salts are, after all, more liable t o undergo intra-molecular changes than are the carbinols, and are thus less suitablefor the determination of true optical values.This idea is supported by a study of a series of active esters ofthe type CH3*CH(O*CO-R)-R’, which were prepared from thecorresponding active carbinols.9 The examination of numerousexamples of these esters has included the effect of temperature,change of concentration, the influence of solvents, and the deter-mination of the rotations in light of different wave-lengths.Themain conclusions drawn are: (1) alterations in rotatory powerobserved owing to change of temperature or the conditions of solu-tion, are due to alteration in the nature or complexity of the activemolecules, and (2) the anomalous dispersion exhibited by someT.M. Lowry, T., 1914, 105, 81.5 T. M. Lowr!, R. H. Picknrd, and J. Kenyon, ibid., 94.* R. H. Pickard and J. Krnyon, itid., 1115.Ibid., 830, 2262 ; J. Kenyon. ibid., 2226ORGANIC CHEMIS‘l’ltY. 67esters of apparently simple structure points t o the occurrence ofintramolecular changes in the .ester group. This supposed changeapparently cannot involve alteration in mass, and is probably dis-tinct from association or polymerisation. One suggestion offered isthat the ester group may, in common with the carboxyl group,react either asorthe proportion of each forni varying according t o the conditions,although the authors are careful to state that they obtained nodirect evidence indicating the presence of such isomerides in theesters examined by them.The results of this and other re-searches10 afford strong support t o the recent expansion of the ideathat anomalous dispersion is due t o the simultaneous presence oftwo interconvertible active compounds possessing different dis-persive powers.11Evidently a new feature of complexity must now be recognisedin quantitative investigations connected with optical activity.Optically pure liquids must be examined under conditions whenthey are actually homogeneous, and this, no doubt, will play aprominent part in the future development of this importantsubject.Considering the manif old difficulties which surround this problem,it is surprising to find that, even in compounds which containseveral asymmetric systems, some empirical quantitative generalisa-tions have been established.Among these may be mentionedHudson’s rule regarding molecular rotation in the sugar group.12Although this generalisation is based on principles which are notgenerally accepted, it has been shown to apply exactly in the caseof the a- and P-forms of monomethyl glucose.13 As these iso-merides showed suspended mutarotation, i t was possible to deter-mine the maximum optical values dii-ectly, the results obtainedbeing in agreement, within the limits of experimental error, withthe values calculated by Hudson’s method.It is possible that the whole subject of optical activity may, inthe near future, undergo startling changes, and much interest willbe taken in the results recently contributed by Erlenmeyer andhis collaborators on asymmetric syntheses effected by means ofThe well-known experiments quoted by asymmetric induction.”G.W. C!ough, T., 1914, 105, 49.11 H. E. Arrnstroiig and E. E. Walker, PTOC. Zhy. SOC., 1913, [ A ] , 88, 388 ; A.,1913, ii, 543.Ann. Xeport, 1909, 124 ; 1913, 79.13 J. c‘. Irvine and T. P. Hogg, T., 1914,105, 1386.F 68 ANNUAL REPORTS ON THE PROGRESS OF CHEMJSTRY.Marckwald, in which active a-metliylbutyric acid was obtained bydecomposing the brucine hydrogen salt of inetliylethylmalonic acid,have been repeated under conditions which exclude the separationof the salt in fractions. The ultimate product obtained neverthe-less displayed optical activity, and this result is interpreted byErlenmeyer as a proof that two isomeric methylethylmalonic acidsexist, and that the formation of an active inethylbutyric acid isdue to induction.Further, & or Z-tartaric acid may take theplace of brucine in a parallel reaction, giving rise similarly to theZ- or d-form of methylbutyric acid.14 Results which are less easilyexplained have, in addition, been reported, as cinnamic acid issaid to have been obtained in d- and Z-forms by heating with theactive tartaric acids.15 According t o Erlenmeyer, cinnamic acidis to be regarded as C,H,*CHL-C'€IL*CO,H, where L representsan unoccupied valency power the existence of which permits of theformation of non-superimposable forms.Scrutiny of the experi-mental details leaves a certain amount' of doubt as t o whether theimportant claim has been justified that a compound devoid ofan asymmetric atom (or its equivalent) has actually been obtainedin active forms, or, a t all events, has been cause:!, to acquire opticalactivity. At the same time, the idea does not rest on the evidenceof a single example, as benzaldehyde has also been added to thelist of compounds which can be made to display induced activity.16The development of this work will doubtless be closely watched.Tau tomerism.Although in recent Reports considerable attention has beengiven to the consideration of tautomeric change, and the subjecthas also been reviewed in the latest Presidential Address,l7 it isdifficult to resist the temptation t o refer t o some new observationswhich seem worthy of mention.The suggestion that ozone will prove t o be a useful reagent inthe study of tautomerides has been considerably strengthened inthe course of the past year, and in an important paper18 a numberof fresh examples illustrating its efficiency are quoted.I n addi-tion, evidence has now been accumulated to show that ozone doesnot act catalytically in affecting the keto-enol change, and that itsreaction is confined strictly to enolic forms. Now that these pointshave been established, the method will doubtless be extensivelyl4 E. Erlenmeyer and F. Landsberger, Biochem. Zeitsch., 1914, 64, 366 ; A., i, 920.l6 E.Erlenmeyer, G. Hilgendorff, and F. Landsberger, ibid., 296 ; A . , i, 965.l6 Ibid., 382 ; A . , i, 967.l7 W. H. Perkin, jun., T., 1914, 105, 1176.l8 J. Scheiber and P. Herold, AnnaZen, 1914, 405, 295 ; A, i, 926ORGANIC CHEMISTRY. 69used, particularly as the experimental procedure appears to becomparatively simple. I n the paper referred to, the results quotedare for the most part perfectly normal and in agreement withthose arrived at by other methods, but one or two special casesare considered. Thus, the mixture of ozonides obtained fromoxalacetone gave, on decomposition, a variety of products thenature of which proves that the parent diketone exists in three un-saturated forms, two being mono-enolic and the third di-enolic :OH*CMe:C:C( ON)-CO,E t.COMe*CIX:C(OH)*CO,Et, O€I*C1Clle:CH*CO*CO,Et,111 view of this result, aiitl others of a siniilar nature, i t is notsurprising that the applicatioii of Meyer's volumetric method todiketones frequently gives figures poiiiting to an eiiol content ofnearly 100 per cent., and, in the case of oxalacetone, the resultis above this maximum.Nevertheless, Meyer's process continues to give valuable results,and, in some instances, adequate explanations have been fortli-coming to account for the discrepancies occasionally encouuteredin comparing the values obtained by this method with thoseindicated by purely physical processes. Thus, on redeterminingthe refractive indices of pure (ketonic) acetoacetic ester, and alsoof the equilibrium mixture,lg values have been obtained which arein agreement with the result, previously arrived a t by the titrationmethod. The conclusion is again drawn that, in the liquid state,the equilibrium mixture contains from 7 to 7.4 per cent.of theeiiolic form, and as this resuIt is considerably higher than thatquoted by Knorr, it is boldly suggested that the latter workerfailed to determine the refractive index of the keto-ester untili t had undergone partial enolisation.An interesting application of the bromine-titration process isindicated in a paper, where it is shown that. the desmotropicchanges exhibited by nitro-compounds can be followed quantita-tively, just as in the case o l keto-enols. Phenyldinitromethaneproved a specially suitable test substance in this respect, and asthe mi-form is more soluble than the true nitro-derivative, theeffect of water and the simple alcohols is the reverse of thatoccasioned with tautomeric ketones.20 Current ideas regarding thestructure of isonitro-compounds are, however, still somewhat un-settled.21The idea that, in the case of aldehydes, substitution by meansof halogens is preceded by a change comparable with the keto-enoll 9 I<.H. Meyer znd F. C. Willson, Bcr., 1914, 47, 837 ; A , , i, 483.2o K. H. Meyer and P. Wertlieiirier, ibiJ., 2374 ; A . , i, 1061.21 S. S. Xanietkin, J. Riw. Phys. G'hem. Soc., 1913, 45, 1414 ; A., 1913, i, 129770 ANNUAL REPORTS ON THE PROGRESS OF CHEMISTRY.transformation has been realised in the course of a dynamica.1study of the action of bromine or iodine! on acetaldehyde.22 Itwas found that, in the absence of acids, bromine acts entirely asan oxidising agent, but that oxidation and substitution proceedsimultaneously in strongly acid solutions.The fact that thevelocity of the reaction is independent of the concentration andnature of the halogen used may be accounted for on the supposi-tion that three distinct changes are involved, the first of whichis enolisation :CH,*CQH 0 --+ CH,:C<EH -> CH,X*CX<gH + CH,X*C<E<Y A < / \ /Y Enolisation. Addition. Loss of halogen hydride.It may be mentioned that, according to, the most recent views,23the formation of iodoform from acetaldehyde is another reactionthat proceeds only with the eriolic form, and i t will be interestingto observe how far the tendency of aldehydes to react in thismanner is of general application.Hydrocarbons.With the exception of investigations on the mechanism of com-bustion, few.publications of the past year have dealt with thechemistry of saturated hydrocarbons. At the present time, mostresearches on unsaturated hydrocarbons lead ultimately to thecaoutchouc problem, with the result that it is often a matter ofsome difficulty t o disentangle from the mass of published workresults which are of general interest and importance. Severalrecent papers have, however, dealt with the formation of poly-merides from hydrocarbons, and a number of fresh contributionshave been made to this long-standing problem.As an example of such work, attention may be directed to acareful study of tetramethylallene, CMe,:C:CMe,, which appearsto have been obtained pure for the first time.24 It is shown thaton polymerisation under the simplest possible conditions the hydro-carbon is converted almost completely into the dimeride, the sbruc-ture of which, in view of its optical properties, is probablyCMe,*$!:CMe2 ICMe2-C:CMe2'On the other hand, more drastic methods result in the forma-tion of acetylenic derivatives, such as CHMe,-CHMe*Ci CH, andthe alteration thus occasioned in the carbon chain points to a con-z2 H.M. DBWSOII, D. Burton, and H. Ark, T., 1914, 105, 1275.28 A. Pieroni slid E. Tonnioli, Gazzetta, 1913, 43, ii, 620 ; A ., i, 6.24 B. K. blereshkovski, J. Ricss. Phys. Chem. SOC., 1913, 45, 1940; A , , i, 369ORGBNlC CHEMISTRY. 71siderable tension in the structure CR,:C:CR,. A general prin-ciple seems to be involved here, as comparison of the velocities ofpolymerisation of allene, di-, tri-, and t,etra-methylallene, provesthat the stability of the compounds diminishes steadily with in-creasing substitution.25 The limiting case is thus reached in tetra-methylallene, where all the replaceable hydrogen atoms aresubstituted by alkyl groups.Halogen derivatives of acetylene are now being used extensively,and in another section of the Report reference is made to asynthetical method of preparing unsaturated glycols which dependson the use of magnesium acetylene haloids.26 Two types of suchreagents are available, namely, CII iC*MgX and MgX-CIC*MgX,and the former, on reaction with ketones, give rise to unsaturatedtertiary alcohols. A greater range of reactions is afforded by theuse of dimagnesium acetylene dibromide, which, naturally, reactswith ketones and aldehydes to give ditertiary y-glycols and di-secondary y-glycols respectJvely.27 As a side-issue of the generalscheme of research, it may be remarked that the results of thisseries of investigations point conclusively to the symmetrical struc-ture for the halogen derivatives of acetylene.2* This point wasreferred to last year, when it was shown that Nef's views as tothe structure of these derivatives were open to criticism.Caoutchouc.--As was t o be expected, the list of patented pro-cesses for the preparation and polymerisation of hydrocarbonsrelated to caoutchouc bas been materially increased during thepast year.The methods employed continue to show greatdiversity, and there is a natural tendency to range far and widein search of polymerising catalysts, but, apart from this aspect ofthe general problem, i t is evident that our ideas regarding thestructure of the caoutchouc complex must, in the meantime, remainsomewhat less decided than was the case a year ago. Before deal-ing with new experimental facts, mention may perhaps be madeof an interesting historical account of the earlier work on thesynthesis of caoutchouc which will be useful in teaching, providedless prominence is given t o the controversial aspects of the~ubject.~QAs is well known, the claim made by Harries that the caout-choucs prepared from isoprene, by ,autopolymerisation, or by theaction of acetic acid, are identical with each other and with the"5 S.V. Lebedev and B. R. Mereshkovski, J. Bass. Phys. Chcm. SOC., 1913, 45,a6 5. I. Iooitsch, ibid., 1902, 34, 239 ; ' 4 . , i, 393.17 Ibid., 242 ; A . , i, 405.29 F. J. Poiid, J. Amer, C'hem. SOC., 1914, 36, 165 ; A., i, 194.1249 ; A . , 1513, i, 1285.Ibid., 1904, 36, 1545 ; A., i, 37372 ANNUAL REPORTS ON THE PROGRESS OF CHEMISTRY.natural product, depends on the identity OF the respective ozonides.A similar comparison of the caoutchoucs derived from isoprene bythe action of sodiuin o r of peroxides leads, however, to the con-clusion that in each case the products are non-honiogeneou~.~~From the fact that the ozonides thus obtained decompose to givesuccinic acid, acetonylacetone, lzvulinaldehyde, and lzevulic acid, itis deduced that the caoutchoucs must contain, not ollly1 : 5-dimethyl-Al ::’-cyclo-octadiene, but also the corresponding1 : 6-dimethyl isomeride :1 $JH,*CMe:CH.$JH,C H, C 31 e : CH CIS, 2’I n order to explain this result, it is suggested that the differentpolymerides arise, respectively, from the symmetrical and the un-symmetrical condensation of two molecules of isoprene.The state-ment is also made that the caoutchouc prepared by autopolymerisa-tion of isoprene is likewise’ a mixture, and is similar t o “sodiumcao~tchouc,~~ in contrast t o which the natural product is derivedentirely from 1 : 5-dimethyl-A1 :5-cycZo-octadiene.According to theresults of this research, the whole question of the identity or non-identity (and also of the uniformity) of artificial and naturalcaoutchoucs is once more thrown open. To this Harries has re-plied, admitting the irregularity with which the polymerisation ofisoprene proceeds, but suggesting that the source of the above con-flicting results may be traced to the presence of impurities in theisoprene used by Steimmig.31 The suggestion has been followed bywhat might, with advantage, have preceded this discussion, namely,a redetermination and standardisation of the physical constants ofthe parent hydrocarbon.32 It may, in passing, be noted that thevalues now quoted for the density and refractive index of isopreneare in fair agreement with those obtained by other workers.33It is necessary to state that the most direct evidence in favourof the idea that the caoutchouc complex consists of an aggregateof eight carbon rings has now been withdrawn.34 As explained inlast year’s Report, it is possible to degrade a “regenerated caout-chouc ” by conversion into the ozonide, and treatment of the latterwith water.The diketone thus obtained as the essential productwas then considered t o be cyclo-octane-1 : 5-dione7 but repetition ofthe work on a larger scale has shown that such is not the case.The degradation, in fact, results in the formation of heptane-P<-30 G.Steimmig, Ber., 1914, 47, 350 ; A . , i, 307.31 C. Harries, ibid., 573 ; A , , i, 422.9’L Ibid., 1999 ; A . , i, 917.y4 C. Harries, B e y . , 1914, 47, 784 ; A., i, 386.S. V. Leberlev and B. K. Mereshkovski, Zoc. citORGANIC CHEMISTRY. 73dione, CH,-CO* [CH,],*CO*C)H,, which readily undergoes dehydra-tion, and this alteration in composition leads to a fortuitous agree-ment with the analytical figures required for C8H,,0,. Of thenature and identity of this diketone there can be no further doubt,as the same compound has been obtained in the course of entirelydifferent work by the oxidation of heptan-<-01-P-one by chromicacid.35Although thus deprived of what appeared t o be a most con-vincing experimental proof, the idea that caoutchouc is derivedfrom an eight-membered ring is still strongly supported by othcrresults.A t the same time, somewhat different views are expressedin the course of an elaborate research36 in which the mechanismof polymerisation in olefi ne hydrocarbons is discussed. The generaliiiethod adopted in the research now under review was t ostudy the two dimerides of isoprene, and of related hydrocarbons,as a first stage in the more complex polymerisation. The resultsobtained leave the impression that rigid views on the whole ques-tion are not yet justified.Acids and Related Compounds.As the acids of the aliphatic series have in the past been thesubject of prolonged and careful investigation, it is not surprisingto find that a stage has now been reached when few discoveries offar-reaching importance have t o be recorded. Nuch of the re-search now being done on the simple aliphatic acids has a directphysical bearing, and, as a result, improved methods of purifica-tion and revised physical constants are frequently reported in theliter a ture.Another type of physico-chemical research, in which aliphaticacids are involved, may be mentioned a t this stage, as, although thesubject is still incompletely developed, i t will, no doubt, be followedwith much interest.The general inquiry as to the nature of soapsolutions has been specially prominent of late, and some results ofan unexpected nature have now been established. For example, theconductivities of thO potassium salts of fatty acids have been deter-mined for all compounds, with an even number of carbon atoms,from acetic acid up to stearic acid.37 The values found are notablyhigh, and, taken in conjunction with previous work on this sub-ject, i t is evident that high conductivity is a general feature ofthese salts even when the solutions approach the colloidal con-dition.Moreover, a determination of the degree of hydrolysis and35 1%. G. Farglier and W. H. Peikin, juii., T'., 1914, 105, 1353..'G 9. Y. Lebedev and B. li. Mereshkovski, Zoa. c i t .13. M. Bunbury an11 H. E. Martin, T., 1914, 105, 41774 ANNUAL REPORTS ON THE PROGRESS OF CHEMISTRY.the true alkalinity of map solutions shows that the amount offree alkali present in such systems is insufficient to account forthe conductivity vaIues.38 I n order to explain the behaviour ofthese solutions as electrolytes, the suggestion is made that thismay be due to the presence of highly charged aggregates whichare not ions in the ordinary sense of the term, but for which theexpression " colloidal ion " might reasonably be applied.Work ofthis nature offers many possibilities for theoretical expansion, butmust present many experimental difficulties.There is no other problem connected with unsubstituted mono-basic acids which calls for special attention, but, on the otherhand, the succinic acid series continues t o prove an attractivefield of inquiry. For a considerable time it has been evidentthat the analogy between phthalyl chloride and succinylchloride is not well maintained, and that, generally speaking, thereactions of the latter compound are by no means in agreementwith an unsymmetrical cyclic structure.The action of ammoniaon the acid chloride certainly gives very small yields of succin-amide, but i t is now shown that, in the reaction with methyl-amine, the symmetrical product is formed to the extent of 25 percent., and ultimately, in the case of aniline, ths reaction of asimilar type is quantitative.39 I n all probability, the differentresult obtained in thme reactions is due to the superior effect ofammonia in removing the elements of hydrogen chloride from theint'ermediate compounds formed in each case. This is illustratedin the original paper by means of a structural scheme which dis-penses with the necessity of modifying the symmetrical formulafor succinyl chloride.The adoption of a similar scheme wouldserve to explain the observation4O that succinyl chloride, in its re-action with zinc alkyl iodides, gives rise only to products of the>o. c €3,--COCH,*CR,lactonic type, IOn the whole, the accumulated evidence to which reference wasmade in last year's Report is in favour of the normal formula forthe chlorides of all acids of the succinic type. Certainly in thecase of d-dimethoxysuccinyl chloride the optical behaviour of thecompound indicates that it possesses an open-chain structure, andthat, even under the influence of powerful catalysts, it shows notendency to undergo tautomeric change into a lactonic modifica-tion.41 An additional example which emphasises this distinction* J.W. 1lcB;sin and H. E. Nartin, T., 1914, 105, 957.39 G. F. blorrell, ibid., 1733.40 E. F,. Blake, Compt. rend., 1914, 158, 504 ; A , , i, 384.41 T. Purdieand C. R. Young, T., 1910, 97, 1524ORGANIC CHEMISTRY. 75between phthalyl and succinyl derivatives is furnished by a st,udyof the optically active anilic acids and anils obtained fromd-dimethoxy- and d-diethoxy-succinic acids re~pectively.~? No iso-anils were isolated, and as the activity of dimethoxysuccinaail wasnot affected by reagents which usually induce intramolecularchaiiges, the conc!usion may reasonably be drawn that the com-pound is a definite chemical individual. I n the same paper i t isalso suggested, on theoretical grounds, that Auwers’ views as tothe configuration of the s-dialkylsuccinic acids require alteration,and that the racemic modifications are not the compounds ofhigher, but of lower, melting point.This idea has been confirmedexperimentally, as it has been shown that whereas the dimethyl-succinic acid melting a t 1 9 5 O is irresolvable, the isomeride oflower melting point (127O) is the racemic form, and may beresolved by the agency of triethylenediaminecobaltic bromide.43This result furnishes, incidentally, an example of the somewhatunexpected application of active cobaltammines iil resolutions, andalso indicates the caution which must be exercised in allocatingmeso- and racemic configurations on the basis of melting-poiiitcomparisons.It has long been recogniszd that the replacement of hydrogenatoms by alkyl groups imposes restrictions on many reactions ofderivatives of succinic acid, and this is particularly noticeable inthe formation of aniides from esters.An extreme case is illus-trated by the fact that whereas methyl succinate readily givesan 80 per cent. yield of succinamide, a parallel reaction withmethyl cis-a/3-dimethylsuccinate proceeds with difficulty, andgives only 2 per cent. of the corresponding amide.41 It will beseen that this result is in agreement with Fischer’s view that’ thefirst action of ammonia on an ester is the formation of an un-saturated salt, and lends support to his prediction that a tetra-nietliylsuccinic ester would, in consequence, fail to react withammonia.Meth?/lcarbonato-acids.--Tlie methylcarbonato-derivatives ofhydroxy-acids are evidently of special value in syntheses where itis necessary to exclude the hydroxyl group from reaction, andseveral examples of their use in this respect have recently beendescribed. The characterisation of simple types of these com-pounds is, however, somewhat imperfect, and a detailed accountof metliylcarbonatoacetic acid will be welcomed.45 The acid isreadily acted on by thionyl chloride to give methylcarbonatoacetyl42 C.R Young, T., 1914, 105, 1228.43 A. Werner and M. Basyrin, Ber., 1913, 46, 3229; A., 1913, i, 1302.44 CT. F. Morrell, T., 1914, 105, 2698.45 E. Fischer and H. 0. L. Fischer, Ber., 1914, 47, 768 ; A., i , 38176 ANNUAL REPORTS ON THE PROGRESS OF CHEMISTRY.chloride (I), from w.hich the corresponding methylanilide (11) iseasily obtained :COCl*C'H,*O*CO,Me -+- NMePh*CO*CH,*O-CO2Me.As was to be expected, the successive action of sodium hydroxideand hydrochloric acid on the latter compound resulted in theformation of the methyladide of glycollic acid.I n anothercharacteristic change, methylcarbonatoacetyl chloride was broughtinto reaction with benzene in the presence of aluminium chloride,and, on acidification of the solution, methylcarbonatobenzoyl-carbinol, COPh*CH,*O*CO,Me, was produced. From this, in turn,by the action of alkali, the carbomethoxy-group was expelled, withthe formation of benzoylcarbinol. Similar reactions have beencarried out on the a-methylcarbonatopropionic acid derived fromlact'ic acid, and consideration of the whole series of changes givesthe impression that this new method of temporarily masking thereactive hydroxyl group is extremely important. The use ofacetyl derivatives for the purpose mentioned above is lessefficacious, as, in addition to the ease with which the sub-stituting - groups are removed by hydrolytic reagents, the tendencyof these compounds to lose acetic acid is frequently very great.Occasionally this inst ability leads to interesting results, as, forexample, in the conversion of diacetyltartaric anhydride intocarbon s u b ~ x i d e .~ ~ Apparently this reaction involves the elimina-tion of two molecules of acetic acid, thus giving rise to acetylene-dicarboxylic anhydride, which, in turn, is converted into carbonsuboxide through the intermediate formation of P-oxypropiolo-lactone :(1.) (11.)AcO*C'H*COA ~ O ~ H - C O c*co c*coEsters.-As most of the synthetical work involving the use ofesters is described under various other headings, there is littlenecessity t o devote a special section to these compounds, particn-larly as few novel or unusual results have been noted in the periodunder review.Esters still find manif old applications in condensa-tion reactions, their halogen derivatives are usefully employed insyntheses involving organo-metallic compounds, and, as in the past,they have been the subject of careful physical examination.Attention may be drawn t o a st'udy of the hydrolysis of methylacetate in which hydrogen chloride, in widely varying concentra-tion, was used as a catalyst.47 I n this specific case i t has beenfound that there is no tendency for the temperature-coefficient of46 E.Ott, Ber., 1914, 47, 2385; A . , i, 1048.47 A. Lainble and V7. 0. M. Lewis, Y'., 1914, 105, 2330ORGANIC CHEMISTRY. 7 7;L strongly catalysed reaction to be less than that of a weaklycatalysed reaction. Evidently, if tlie effect of a catalyst is merelyto increase the ratio of active to unactivated molecules, thetemperature-coefficient should diminish when the number of activemolecules is increased. The results of this investigation certainlylend support to the views expressed by Marcelin that the effect oftemperature in accelerating reactions depends on an increase inthe internal energy of the reacting molecules, and does not involveany distinction between active and inactive molecules. Thetheoretical discussion is extended t o a consideration of catalysis asa radiation effect, and thus falls within the compass of anothersection of the Reports.Lactones.-Probably the most comprehensive study of simplelactones which is to be reported is described by Nef in connexionwith the examination of acids allied to the sugars.This is, how-ever, dealt with later under carbohydrates, and, in the meantime,attention may be directed t o various extensions of earlier work onthe reactions of P-lactones.As a rule, complex changes ensue when a P-lactone is heated,and the formation of ketens in such reactions has been traced inan examination 48 of the decomposition undergone by the esters ofcertain lactonic acids.By the interaction of methyl iodide andthe /3-lactone of silver /3-liydroxyisopropylmalonate, the esterobtained is, curiously enough, not a derivative of the parent acid,but of the P-lactone of /3-hydroxy-u-methylisopropylmalonic acid :$)-QMe2 ?-9Me,C 0- CH C0,Ag --+ C O ~ C M ~ * C O ~ M ~This irregular reaction resembles in many ways the action of alkyliodides on the silver salts of hydroxy-acids, where, in addition t othe normal ester, the corresponding alkyloxy-ester is invariablyformed to some extent. The ester formulated above is remark-ably stable,, but is decomposed when heated in an inert atmo-sphere to give a notable yield of dimethylketen, together withacetone and carbon dioxide :0-CMe,... I I i.......-~.-.~..--.._.___._._..... ._ -+ Me2CO-t CO, + Me,C:CO. ~ : o * u M ~ & o , ~ ~ M ~ IThe reaction appears t o involve molecular rupture in the manherindicated by the dotted lines, followed by the transference of amethyl group to the keten residue, and, although this may appeara somewhat empirical explanation of the change, it is supportedby the fact that bromomethylketen, CBrMe:CO, has been obtainedin a parallel reaction from the corresponding a-bromo-lactonic ester.48 E. Ott, Annnlea, 1913, 401, 159 ; A., 1913, i, 130278 AKNUAL REPORTS ON THE PROGRESS OF CHEMISTRY.A Zdel~ydes and Ketones.-The preparation of aldehydes andketones by methods depending on catalytic agency still continuesto be prominent, but there is little under this heading t o add t othe account given in 1913, except that manganons oxide has bee11used with considerable success as a catalyst in thesc r e a ~ t i o n s .~ ~The reagent is claimed to possess several advantages over titanicoxide in this type of reaction, and is possibly more uniform in itseffect than thorium oxide, which sometimes gives irregular results.During the past few years, the preparation and study of unstablealdehydes have been greatly extended, and a new syntheticalmethod for the preparation of certain types of such compounds hast o be ncted. As is well known, the methods hitherto in use forthe preparation of glyoxals are n o t only experimentally difficult,but are frequently uncert'ain in their results.The following seriesof reactions, however, appear to be general in their application,and to proceed smoothly.50Ethyl diethoxyacetate may be condensed with ethyl acetate bythe agency of sodium:CH(OEt).2* C0,Et + CH( OEt),*CO*CH,*CO,Et.The product (or its homologues) is then hydrolysed by dilute acids,giving, in the first instance, an acetal, and finally the correspond-ing alkylglyoxal, CH(OEt),*CO*R ++ CH0.CO.R. The proper-ties of a number of new glyoxals are now described, and, in future,these compounds will be more readily available. It is significantthat most of the new representatives display the yellow-green colourwhich appears to be characteristic of all true glyoxals.As wasto be expected, the compounds are readily converted into thecorresponding hydroxy-acids, and it is to be noted that the enzyme,glyoxalase, effects the same change, but gives optically active pro-ducts. This is an important observation, the significance of whichwill be obvious.The preparation of aldehydes by direct oxidation of the corre-sponding alcohol is generally a difficult operation, and in this con-nexion it is well to draw attention to a paper in which the pre-paration of pure glyceraldehyde from glycerol is described.51 Theoxidation was effected by means of Fenton's reagent, and the pro-duct, isolated through the intermediate f ormation of the diethyl-acetal, was obtained in the crystalline state.These reactions have,of course, already been carried out, but the special feature of thepaper, which should be consulted by those who have occasion toemploy hydrogen peroxide in similar reactions, is the precise ex-49 P. Sabatier and A. Mailhr, Cornpl. rend., 1914, 158, 830 ; A , , i, 547.50 H. D. Dakin and H. W. Dudley, T., 1914, 105, 2453.51 E. J. Witzemann, J. Amer. Chena. SOC., 1914, 36, 2223 ; A., i, 1165ORGANIC CHEMISTRY. 79perimental conditions which are established, both for the use ofFenton’s reagent and for the production of acetals.Polyhyclric Alcohols and their Derivatives.Few methods for effecting the apparently simple change ofa dibromo-compound of the type of ethylene dibromide into thecorresponding glycol are satisfactory so far as yields are con-cerned.Even the preliminary conversion of the halogen com-pounds into acetates is, in many cases, no more effective than thedirect process, and unsaturated derivatives of the type of vinylbromide are frequently formed, according to the schemeCR2Br*CHRBr -+ CR,:CRBr.It has been shown in a careful study of this process that the useof silver acetate as a reagent for decomposing dibromides is lessopen to objection in this respect than is that of potassium acetate.62Further, the ease with which diacetates are formed from dibromidesis less in higher than in lower homologues, and is diminished whenthe two bromine atoms are in spatial proximity. This a p e s withthe experience of other workers who have been engaged on similartopics, and a number of striking experimental facts are recordedby Franke,53 who has shown that dibromides of the typeCH2Br*CRR’*CH2Brare notably stable towards both potassium cyanide and silveracetate.On the other hand, the above reagents react readily withdibromides constituted according to the formulaCR2Br* CH2*CR2Br,and the suggestion is made that the reaction takes place in twostages, involving : (1) the formation of an unsaturated derivative,and (2) the addition of hydrogen cyanide or acetic acid. Both thepapers now quoted may be consulted with advantage by all whoare engaged with work involving dibromides, which can only beobtained in small quantity, and the suggeetion that sodiumethoxide in dilute alcoholic solution is the most effective reagentfor the hydrolysis of diacetates will doubtless find many appli-cations .Acetylenic glycols offer many possibilities for attractive research,and a representative series of these compounds has been preparedduring the past ten years through the agency of magnesium-acetylene derivatives.A simple example may be quoted in illus-tration of the pr0cess.5~ Dimagnesium acetylene dibrornide reactswith acetaldehyde to give the additive compound,MgBr*O*CHMe*CiC*CHMe*O*NgBr,52 E. G. Bainbridge, T., 1914, 105, 2291.53 A. Franke, Monalsh , 1913, 34, 1893 ; A., i, 7.54 S . I. Iocitsch, J. Russ. Phys. Chern. Sac., 1903, 35, 430 ; A . , i, 37580 ANNUAL REPORTS ON THE PROGRESS OF CHEMISTRY.from which, on decomposition with water, the corresponding un-saturated glycol, O€I*CHMe*CiC-CHMe*OH, is produced.Thereaction, which is applicable t o substituted aldehydes and also toaliphatic or cyclic ketones,55 seems perfectly general, and has beenapplied to a large number of cases. Norw that improved methodsfor tlie preparation of unsaturated glycols are available, the ex-amination of these compounds is being carried out in considerabledetail. 56Glycerol.-When critical research is brought t o bear on reactionsfamiliar t o the lecture-room or laboratory, the result is frequentlyt o unfold new difficulties and t o complicate the task of teacher andstudent alike. The opposite result, and the reduction of importantreactions to simple terms, is thus doubly welcome.The interaction of glycerol and oxalic acid under varying con-ditions has now received a new interpretation, which will be gener-ally accepted.57 Both an acid and a normal oxalate are produced,and the decomposition of these esters gives rise t o all the knownproducts of the reaction.I n the following structural outline,which illustrates most of the changes involved, the double arrowsindicate the stages in which oxalic acid plays a part in the reaction,but the original paper should be consulted for the evidence uponwhich the scheme is based:Glycerol.FH,*O*CO*CO,H YH2*0*C3*H 7 H2*O*CO*C0,HFH*OH -+ YH*OH +GO, I$ ?H*OI-I +H*CO,HCH2*OH CH,*OH CH,*OHAcid oxalate. Monoformin.ICH +2CO,CH,*OH \ 7H2*O*70 SH2Normal oxalate. ' '-.. ~IT--O*CO -+ QH +3C02CH, 0.CO C0,H CH,*O*CO-HAlly1 formate.65 S. I. Iocitsch, J. Ruas. Phys. Chcm. SOC., 1906, 38, 656 ; A . , i, 375.56 R. Lespieau, Compt. rend., 1914, 158, 707 ; A . , i, 476; G. Dupont, ibicJ., 714 ;57 I?. D. Chattaway, T., 1914, 105, 151.A., i, 530ORGANIC CHEMISTRY. 81I n last year's Report, reference was made to one aspect of thecontroversy, which has been maintained for many years, as t o thenature and constitution of tlie glycerylphosphates, and the opinioiiwas expressed that the probleni should be attacked by entirelydifferent methods. During the current year an important paper 58has appeared, in which the cognate literature since 1903 is care-fully reviewed, and, in addition, the fresh experimental facts re-corded have cleared up a number of doubtful points.The simplestpossible glycerylphosphoric acid should exist in two forms,YH,* 0 PO( OH), FH2*OHFH*OH iLtl(f FH*O*PO(OH),CH,*OH CH,*OR(1. ) (11.1and only the a-isomeride should be capable of resolution into activemodifications. Previous attempts t o synthesise the isomerides in astate of purity have only been partly successful, as the tendencyof these esters to form polyglyceryl derivatives is very great. Eventlie identification of the glycerylpliosphoric acid obtained by thehydrolysis of lecithin as an active modification corresponding withformula I is still doubtful. It, has now been shown that theaction of phosphoryl chloride on a-dichlorohydrin is somewhat lesscomplex than was a t one time imagined. The product undergoeshydrolysis in two stages, giving, in succession, bis-s-dichloroiso-propylphosphoric acid and p-glycerylphosphoric acid.As calciumhydroxide and sodium carbonate were employed as the hydrolyticagents, the compounds actually isolated had the followingstructure :(a-Form . ) (B- Form. )CH2C!1 0 CH2C1 CH,*OH(:H.o--Y-o~H -+ C'H-O*PO(ONa), .AH,Cl h? bH,CI UH,.OHThe sodium P-glycerylphosphate proved to be identical with thecompound formed by the interaction of monosodium phosphate andtwo molecules of glycerol, followed by hydrolysis of the resultingdiglyceryl ester.For the preparation of the corresponding a-glycerylphosphoricacid, a new and apparently simpler process was devised in thata-nionochlorohydrin was found t o react readily with trisodiuinphosphate, and from the products of this reaction salts of a definitea-glycerylphosphoric acid were obtained :I I1 IIHO*CH,=CH(OH)*CH,CI + (NaO),PO -+HO*CH,-CH(OH) *CH,*O*PO(ONa),.58 H.King and F. L. Pyman, T., 1914, 105, 1238.REP.-VOL. XI. 82 ANNUAL REPORTS ON THE PROGRESS OF CHEMISTRY.Those who have worked in this field will realise the experimentaldifficulties involved, and the advance which has been made, in thischaracterisation of the simplest types of glycerylphosphates.As pointed out last year, considerable attention is now beingdevoted to the study of glycerol derivatives in which the hydroxylgroups are only partly substituted, or are replaced by differentacyl residues. Numerous papers have since appeared which pointto the continuation of this development, and reference may bemade to the attempts, so far unsuccessful, to synthesis0 opticallyactive glycerides of a special type.59 A glyceride of the generalstructure OR*CH,*CH(OR’)*~~2901t2, where R and R2 representdifferent acyl residues and R’ may consist of an acyl group orhydrogen, should exist in active forms, the asymmetry beingdependent on the glycerol component of the molecule.Compoundsof this description should not only be of considerable physiologicalinterest, but might be of service in the study of lipolytic ferments,and thus lead to the identification of the specific enzymes re-sponsible for the degradation of fats in definite stages. Althoughthe research has not, in the meantime, given many positive results,the work has incidentally furnished a description of a number ofoptically active halogen hydrins.For example, Py-dibromopropyl-amine was resolved by d-tartaric acid and converted into the corre-sponding active bromohydrins by the action of nitrous acid. Inany circumstances, the replacement of both bromine atoms by acylgroups would not be likely to proceed smoothly, and the ease withwhich the active bromohydrins are racemised has, so far, limitedthe development of the research, although the fact that the d- andZ-epibromohydrins seem somewhat more stable leads to the hopethat the original object of the work will be achieved.As in the sugar group, the problems affecting the chemistry offats are now being attacked by two distinct methods.I n the onecase, ordinary synthetical or analytical reagents are employed,whilst in the other the synthetic or hydrolytic functions of enzymesare brought to bear. I n this connexion, many features of primaryimportance are indicated in the results of experiments in whichthe synthetic and hydrolytic effects of lipase are compared.Go Boththe rate and extent of the hydrolysis of triolein by means of lipaseare retarded as the amount of water present is increased, and thesame holds true f o r the synthesis of the triglyceride. The factorscontributing to this result differ, however, according as the changeis hydrolytic or synthetic. I n t7he former case, the water preventscontact between the enzyme and the oil, whilst during the synthetic59 E.Abderhalden and E. Eichmald, Be?-.) 1914, 47, 1856 ; A., i, 801.6o R. E. Armstrong and H. W. Gosney, Proc. Roy. Suc., 1914, [B], 88, 176 ;A . , i, 1149ORGANIC CHEMISTRY. 83reaction the water acts directly by removing glycerol in solution.The important observation is made that not only does the hpdro-lysis of triolein by lipase proceed in definite stages, giving a cli-and mono-glyceride, but that the synthetic action likewise shows atendency to be arrested a t a stage when the diglyceride constitutesthe main product.With the exception of investigations connected with biochemicalproblems, comparatively few researches have been concerned withpolyhydric alcohols higher than glycerol. The detailed chemistryof these compounds is, however, by no means exhausted, and muchwork remains to be done in this field.Considering the importanceof compounds allied to sugars which contain a methyl group i nthe terminal position of the carbon chain, the synthesis of a methyltetritol61 will be recognised as a step in the direction of arrivinga t the configuration of the methyl tetroses and pentoses. The com-pound in question was isolated, together w i d the correspondingaldose, by the reduction of dihydroxyvalerolactone. It may beremarked that although the configuration of active polyhydricalcohols is generally established by their relationship to the sugars,i t is possible to obtain the necessary evidence by another method.The condensation of polyhydric alcohols with acetone gives rise toisopropylidene derivatives, the stability of which is dependent onthe configuration of the hydroxyl groups F o r example,i t is possible by carefully regulated hydrolysis t o remove the threeacetone residues from tri-isopropylidenemannitol in definite stages,and by means of a somewhat complex experimental treatment it ispossible to make this result the basis for the allocation to mannitolof the following configuration :OH H H OH OH OHI I I I I IH.C--C--C--- C-C---OHIH d H d H h I!€ €!CIt will be remarked that, in the above structure, definite posi-tions are ascribed to the terminal primary hydroxyl groups, andthere seems every reason to believe that the idea is not onlyjustified, but furnishes an explanation f o r many of the irregularreactions of active alcohols of this type.Thus, to select one illus-tration, it has been shown that mannitol is only capable of formingpenta-ether~,~~ and that it is impossible to extend the alkylationt o one of the terminal hydroxyl groups. This pronounced sterichindrance is consistent with the idea that in mannitol there arethree hydroxyl groups in close spatial proximity.61 R. Gilmour, T., 1914, 105, 73.J. C. Irvitio and Miss B. M. Paterson, ibirl., 898. 63 Ibid., 91584 ANNUAL REPORTS ON THE PROGRESS OF CHEMISTRY.Curb ohydrut cs.The progress made during the year in the study of the simplercarbohydrates has, in many respects, been remarkable, and, in allprobability, will lead to future developments of the utmost import-ance.I n the course of the past ten years the adoption of they-oxidic formula f o r reducing sugars and their glucosidic derivativeshas practically become universal, and there is thus a certain dangerthat, by blind adherence to its use, restrictions may be placedbotlh on synthetical methods and on legitimate speculations as t ostructure. This danger may be avoided if sufficient attention ispaid t o some recent publications,As is well known, Nef has been engaged for many years instudying the effect of simple oxidising media on carbohydrates, andthis inquiry has involved a re-investigation of an old problem-thechanges undergone by reducing sugars in the presence of alkalis.Following the usual method adopted by this author, publicationhas been delayed until the results of several independent investiga-tions were complete, so that it is now possible to view the resultsas a whole.The method has much to commend it, but leaves thereviewer faced with the difficulty of describing, within a limitedcompass, a large field of important work and a mass of experi-mental results. The original paper 64 should therefore be con-sulted, as it is manifestly impossible t o give more than an outlineof the research and to indicate its relationship t'o other work.When an aqueous solution of a hexose is allowed to remain incontact with dilute alkaris, profound changes ensue, and finally anequilibrium is established between the various enolic forms deriv-able from the parent sugar.The enolisation, and consequently thenumber of possible products, is diminished when the concentrationof the alkaline hydroxide is reduced, but in its fundamental prin-ciples t~he reaction is general, although complicated by the forma-tion of resins and polysaccharides. The unsaturated moleculeswhich are formed show a tendency t o undergo fission a t the doublebond, so that, taking glucose as a type, the following productsresult:From glucose aP-dienol+ formaldehyde and arabinose.From glucose Py-dienol+ diose and triose.From glucose y&dienol+ glyceraldehyde.The action of mild oxidising agents on mixtures such as thatdescribed above results in independent oxidation of the coni-ponents with the formation of carbon dioxide, formic and oxalicacids, together with a series of hydroxy-acids.The identificationts J. U. Nef, Annalen, 1914, 403, 204; A., i, 490ORGANIC CHEMISTRY. 85of the latter compounds necessitated their separate preparation,the re-determination of their physical constants, and a study of theconditions under which Obey are converted into lactones. Theresults are somewhat unexpected. d-Mannonic acid, for example,is easily converted into d-mannono-p-lactone,OH CH,* [ C H*OH ],*CH*CH( OH) COwhich rapidly undergoes spontaneous conversion into the parentacid in aqueous solution. The compound thus differs sharply fromthe correspoiidirig y-lactone, which is comparatively stable. Simi-larly, gluconic acid yields both a P- and a y-lactone, and thegeneralisation is claimed that hydroxy-acids corresponding with theformula C,,H,,O,,+ yield both bimolecular a-lactones and uni-molecular p- and y-lactones.It may be remarked that the proper-ties of the alkylated mannonic acids lend no support to the ideathat a-lactones are readily formed.65 Thus, of the two substitutedmaiinonic acids indicated below, only that with the constitutionexpressed by formula I forms a lactone, whilst the remainingisomeride could not be converted into any such derivative:I-.-,-.- IOM~*CH,*CH(O&~~)*CH(OH)*[CH*OM~Y]~*CO~H(1.)OMe*CH,*[CH*OMe],*CH(OH) *CO,H.(1 1.1Although the combined experimental results obtained by Nef areobviously of great importance, some of the theoretical conclusionsarrived at’ seem premature; f o r example, he puts forward the ideathat the a- or P-lactone structure may be applied t o glucosides,and, although this is quite admissable, it is most unlikely that theisomerism between the two crystalline methylglucosides can bereferred t o a difference in the mode of linkage of the ring-formingoxygen atom.According to existing views, these glucosides may berepresented by the formula: A and B, shown below, whilst, in Nef’sA . B. C. D.HC-OMe RleO*CH CH*OMe /CH*OMe~ H ~ O H ~ H - O H /&H.OH 0 AH-OH\dlH M H \&H ~ H ~ O Ko( 6H.OH ‘&I W, bHoOH .< LH~OH /I I ICH*OH CH-OH ~ H ~ O H CH*OHI I ICH,*OH\ / \ J YCH,*OH CH,*OH &H;OHIa-Methyl- a-Form B-Forniglucoside. P-PONIl. (stable). (unstable).65 J. C. Irvine and Miss B.M. Paterson, Zoc. cit86 ANNUAL REPORTS ON THE PROGRESS OF CHEMlSTRY.opinion, the isomezism does not depend on the position of themethyl group, but, as indicated in formulae C and D, on the exist-ence of different ring-structures in the molecule.These conclusions have been chalIenged by Fischer,66 who tabu-lates powerful arguments in favour of the older formulae and a tthe same time describes a new isomeric - form of methylglucoside(termed by him ‘‘ y-methylglucoside ”), which is characterised bythe extreme ease with which it is hydrolysed. For this reasonformula D may possibly be applicable to the new type of glucoside,although the exact linking is still uncertain. It may be remarkedthat, in addition to the evidence quoted by Fischer, the olderformulae for the alkylglucosides are also supported by recent deter-minations of their dissociation constants.67The present position of this subject is full of possibilities.Fischer’s discovery shows thatl recognition must now be given t o thepossible occurrence of a new mode of linking, both in glucosidesand in disacchasides.The structure of many important com-pounds, including the fructosides generally and sucrose in parti-cular, is involved. Although Fischer’s results are so far incom-plete, they can be verified by the experiences of the writer of thisReport, as y-methylglucoside, in the form of its methylated deriv-atives, has been recently encountered in attempts to prepare tri-methyl glucose from glucosemonoacetone.Another question involved in this‘ new development is the inadequacy of the existing system of nomenclature t o express thestructure of sugar derivatives.The expressions a and /3, asapplied to sugars and glucosides, refer t o definite stereoisomericforms, and their use has become standardised. The same alpha-betical system is, however, also used t o indicate the consecutivecarbon atoms of the sugar chain, and the results described aboveadd considerably to the consequent confusion. Fischer’s y-methyl-glucoside may be described as an a-, /3-, ti-, o r (-glucoside in thatthe oxygen &tom of the ring may be connected t o any one of thesecarbon atoms of the chain, whilst the compounds generally knownas a- and P-methylglucosides may, f o r similar reasons, be termedy-methylglucosides.These are by no means the only exampleswhich point to the necessity for radical changes in the presentusage.The fact that the y-oxidic linking present in on0 sugar derivativeneed not of necessity persist during the formation of other deriv-atives is emphasised by the results lately obtained in connexionwith the compound known as glucal. I n last year’s Report refar-66 E. Fischer, Ber., 1914, 47, 1980.67 L. Michaelis, ibid., 1913, 46, 3683 ; A., i, 16ORGANIC CHEMISTRY. 87ence was made to the unexpected course followed by the reductionof acetobromoglucose, and it is to be noted that the provisionalformula then ascribed to the product has now been modified asthe result of further work.68 The unsaturated and reducing pro-perties of glucal disappear when the compound is reduced catalyti-cally, the change involving the addition of two hydrogen atoms.From this result, and on consideration of the properties of thehydroglucal formed, an alternative st,ructure for glucal has beensuggested.This is indicated in the following scheme, which showsthe mutual relationship of these interesting compounds with theacetobromoglucose from which they are prepared :Acetobromoglucose.0 I OA~.CH,.CH(OA~).~H.CH(OA~)*CH(OA~).~H~~ ' Reduction by zinc and acetic acid J.Triacetylglucal.OAc*CH,* CH- UH,* CH (OAc)*C:CH*OA cI 0 IReduction by hydrogenand palladiumIGluc a1 . Hydrolysis\l0 H*CH,*~H*CtI,PCH(OH)*~:~~**~~ +Triacetylhydroglucal.OAc*CH,* &K*CH2*CH(OAc)*CH*CH2*OAc -0- I \H ydroglucal._- --upOH*CH,*bH* CH,* CH(OH)*UH*CH,*OH IThese remarkable and unexpected changes, taken in conjunctionwith Nef's results, will no doubt foster the development of thechemistry of the sugars on less stereotyped lines than has recentlybmn the case.Reactions of the type described above are notconfined t o glucose alone, but are apparently of general application.Thus, lactose 69 and cellobiose70 have been respectively converted68 E. Fischer, Ber., 1914, 47, 196 ; A . , i, 252.69 E. Fischer and G . 0. Curme, ibid., 2047 ; A., i, 931.7O E. Fischer and K. von Fodor, ibid., 2057 ; A . , i, 93288 ANNUAL REPORTS ON THE PROGRESS OF CHEMISTRY.into lactal and cellobial, which resemble glucal in their essentialproperties, and undergo similar hydrogenation t o give saturatedproducts.It is conceivable that many natural compounds, closelyrelated to the sugars, may possibly be derivatives of the glucals,and the suggestion has already been made71 that the “carbo-hydrate” group present in the nucleic acids is in reality related toglucal or an allied compound.A novel method €or the resolution of racemic sugars has alsobeen described, which is based on the formation of active mer-captals by the action of d-amyl mercaptan on inactive aldoses.72As mercaptals are usually easily purified, the process would appearto be experimentally simpler than the methods which depend onthe use of active hydrazine derivatives, but its range of applicationmay be restricted by the fact that the constitution of mercaptalsof this nature is quite unknown.GZucosides.-In last year’s Report, attention was directed torecent advances in the synthesis of glucosides by means of enzymes,and, although considerable progress has been made in the periodnow under review, it is unnecessary t o refer again to this line ofwork in anything like detail.It may be mentioned, however, thatin addition t o the synthesis of several galactosides and other com-pounds of simple type, the process of enzyme synthesis has nowbeen successfully applied to the formation of definite monogluco-sides of glycol and glycer01.7~ These results are mentioned as theyfurnish additional evidence that this method of synthesis can beutilised for the preparation of hydroxy-glucosides which are onlyobtained with exceptional difficulty by ordinary synthetical pro-cesses.The constitution of glucosides and many other structural ques-tions in the sugar group may be solved through the agency ofmethylation, but work of this nature has, in the past, been some-what restricted owing to the expense involved in preparing alkyl-ated sugars by the silver oxide method.A considerable simplifica-tion has, however, been effected, in that, under carefully controlledconditiosns, methyl sulphate and sodium hydroxide may be em-ployed for this purpme.74 I n view of the sensibility of manysugars and glncosides t o the action of acids or alkalis, the use ofthese reagents may seem surprising, but the results already avail-able show that the method is general in its application, whilst thofact that octamethyl sucrose can be prepared in quantity by this71 R Feulgen, Zcitsch.physiol. Chem., 1914, 92, 154 ; A . , i, 1098.E. Vatoeek and V. Vesely, Ber., 1914, 47, 1515 ; A., j, 664.73 E. Bourquelot and 31. Bride], Compt. rcnd., 1913, 157, 1024 ; A., i, 72 ; zbid.,7-l W. N. Haworth, P . , 1914, 30, 293 ; T., 1915, 107, 8.898 ; A , , i, 499 ; ihid., 1219 ; A . , i, 662ORGANIC CHEMISTRY. 89process suggests that the structural problems of the di- and poly-saccharides may now be investigated with reasonable prospects ofsuccess.There is no diminution in the synthetic applications of aceto-bromoglucose, and special reference must be made to the newseries of purine glucosides,75 which have been syntliesised by meansof this popular reagent.The general method of preparationadopted was t o employ the various purines in the form of theirsilver derivatives, and to act on the acetylated glucosides thusproduced wibh ammonia. The compounds isolated are of specialinterest, particularly in view of the fact that they may, in thefuture, be condensed with phosphoric acid, and thus give rise t osynthetic nucleotides. I n subsequent papers 76 the method is shownto be general, both with regard t o the nature of the purine andsugar constituents of t,he glucosides. The outstanding property ofthese compounds is the ease with which they are hydrolysed intotheir components, and this instability, which distinguishes purineglucosides from glucosamine derivatives, is doubtless due t o thefact that, in the former, nitrogen is directly attached to the carbonatom of the sugar chain which functions in the formation of thereducing group.77 Evidence is, in fact, accumuldting which pointsto the idea that the great stability which characterises nitrogenderivatives of the nature of glucosamine is due t o the formationof ring structures when an amino-group is in spatial proximity t othe acidic reducing group. This is emphasised in a publicationdescribing the conversion of glucosamine into mannose,78 and alsoreceives strong support from other work, where i t has been shownthat some of the reactions of a-amino-/3-hydroxy-compounds canonly be explained on the assumption that these substances maybehave as cyclic structures.79Another novel series of glucosides has lately been synthesisedby the interaction of acetobromoglucose and the silver salts ofthiourethanes.80 The I' mustard oil glucosides " obtained afterremoval of the acyl groups were, on further hydrolysis, convertedinto tliioglucose, which was isolated in the form of the silverderivative.Unfortunately, owing to the somewhat indefinite pro-'s E. Fischer and B. Helferich, Bey., 1914, 47, 210 ; A . , i, 333.76 E. Fischer, ibid., 1377 ; A., i, 662 ; E. Fischer anci K. von Fodor, ibid.,77 J. C. Irvinc, R. F. Thomson, and C. S. Garrett, T., 1913, 103, 238.78 .T. C. Irvine and A. Hynd, ibid., 1914, 105, 698.8O W. Schneider, D. Clibben, G .Hullweck, and W. Steibelt, Bey., 1914, 47,1258; A., i, 6 6 9 ; W. Schneider and D. Clibben, ibid., 2218; A., i, 977;W. Sc!ineider and F. Wrede, ibid., 2225 : A . , i, 977.1058 ; A , , i, 741.J. C. Irvine and A. W. Fyfe, ibid., 164290 ANNUAL REPORTS ON THE PROGRESS OF CHEMISTRY.perties of these compounds, their charaderisation is in the mean-time incomplete.Much patient and useful work continues to be expended on tQeisolation of natural glucosides, and the number of galactosidesrecently obtained in this way is st’riking, although there is stillconfusion as to whether these compounds represent cleavageproducts, derived from complexes, or exist in plants or animals inthe preformed state.81 This question has been raised before, andi t ia evident that, in view of the increasing number of galactosederivatives now being described and of their somewhat unusualproperties, the examination of the parent sugar in greater detail ishighly de-sirable.The natural alkaloidal glucosides are also com-pounds which present many difficulties, and it would appear thatthe meagre and confusing references t o these substances to be foundin the literature can be explained by the fact that several differentcompounds of this type have been described under the general nameof ‘’ solanine.” Some of these compounds are exceedingly com-plex,82 and it, is therefore important t o notice that a simple repre-sentative of the class has been isolated from Solanunz a m p s t i -foZium.83 Another interesting observation 84 is the extraction of adibenzoylglucoxylose from the leaves and stems of Daviesia Zatifolia,as this substance represents an entirely new type of natural com-pound.The glucoside is hydrolysed by alkalis with extreme easeto give the non-reducing disaccharide, Cl1H,,O,,, which, in turn,yields both glucose and xylose on complete hydrolysis. The occur-rence of bwo benzoyl groups in a natural compound is most unusual,and further research on the allocation of these substituents t odefinite positions will be awaited with much interest.Tannin.-Although comparatively few new experimental resultsregarding the constitution of tannin have been recently described,it is evident that the views advocated by Fischer are now meetingwith general acceptance. In particular, mention may be made ofthe fact that Nierenstein has now withdrawn his criticisms ofFischer’s structure, and brings forward evidence, based on thebehaviour of tannin towards yeast, that the glucose residue is anessential and combined constituent of the tannin molecule.s5 Thefact that specimens of tannin derived from different sources showvarying specific rotations has, in the past, been quoted in supportof the idea that glucose is present only as an accidental impurity0.Rosenheiiii, Biochem. J., 1913, 7, 604 ; A . , i, 225 ; ibid., 1914, 8 , 110,121 ; A , , i, 706; E. E’ischer, Ber., 1914, 47, 456 ; A . , i, 389.F. Tutin and H. W. B. Clewer, T., 1914, 105, 559.*) G. Odd0 and M. Cesaris, Gnzzelta, 1914, a, ii, 181 ; A ., i, 1173.84 F. B. Power and A. H. Salway., ibid., 767, 1062.85 A. Geake and M. Nierenstein, Ber., 1914, 47, 891 ; A . , i, 567ORGANIC CHEMISTRY. 91in tannin. It has, however, now been shown as the result ofcareful fractional precipitation of commercial tannin that a numberof feebly rotatory isomerides or impurities are present in averagespecimens.86 It may be noted in passing that Fischer’s views admitnot only of the existence of several isomerides, the rotations ofwhich would doubtless lie far apart, but also of the occiirrence ofpartly substituted glucoses in association with tannin.Perhaps the most important recent contribution to the tanninproblem is a critical examination of the constitution of tanninderived from Turkish and from Chinese galls.87 It has already beennoticed that the glucogallic acid isolated by Feist from Turkishgalls is not identical with the P-glucosidogallic acid synthesised fromacebobromoglucose.The opinion hazarded in last year’s report asto the structural difference between these compounds has now beenconfirmed, as, on methylation of glucogallic acid, a non-reducingproduct is obtained, and this, on hydrolysis, yields gallic acidtrimethyl ether. The conclusion is drawn that, in the compoundunder discussion, the glucosidic linking involves the carboxyl group,whereas, in the synthetio isomeride, the coupling of the hexoseand aromatio residues involves a phenolic group. This distinctionin structure is shown below :0-7C,H,(O H),*CO*O*dH*[CH*OH],*CH*CH(OH)*G K,*OII,(I. ) Natural glucogallic “ acid.”CO,EI*C,H,(OH),*O*CH* [CH*OH1,*CH*CH(OH)-CH2*O K.,~ ---()-I(11.) Fischer’s synthetic glucosidogallic acid.A t the same time, although this question of isomerism has prob-ably been definitely settled, the att’empts to confirm formula I bysynthesis are not convincing. The constitution of the glucogallicacids is, of course, a side-issue of the tannin problem, but themethylation process has also been applied by Feist to t’he elucida-tion of the structure of tannins derived from various sources. Theprinciples involved in this method of attack have already beendescribedF8 and, although scrutiny of the experimental detailsgives the impression that the methylation was far from being com-plete, there seems to be no reason why this general method shouldfail to solve the problem, considering the success which has attendedsimilar attempts to methylate cellulose.86 T,. F.Iljin, Ber., 1914, 47, 985 ; A . , 567. *’ K. Feist and H. Ham, Arch. Plzarm., 1913, 251, 468 ; A . , i, 195.s8 J. C. Irvine, Biochem. Zeitsch., 1909, 22, 35792 BNNUAL REPORTS ON THE PROGRESS OF CHEMISTRY.Disacclzarides aiid Yolysaccharides.As usual, the number of papers dealing with the chemistry ofdiswcharides has been very large, but outside of work the mainobject of which was obviously technical or analytical there is com-paratively little t o report.The debatable question as to whether the inversion of sucroseby invertase is reversible or not has been answered by Hudson inthe negative,*g as in the course of an elaborate investigation, carriedo u t with great care and with the aid of specially active invertase,g*the results indicate that the hydrolysis of the disaccharide is com-plete.The synthesis of sucrose by enzyme action is apparentlya reaction which has still to be accomplished, in the laboratory atall events, but, on the other hand, evidence has been obtained9'that dextrorotatory polysaccharides of the nature of glycogen areformed under certain conditions during the alcoholic fermentationof glucose and fructose.Last year i t was noticed that few publications on the complexcarbohydrates were available for review, but 011 this occasion thedifficulty has been reversed, as the contributions under this head-ing have been both numerous and important.Convincing experi-mental evidence has now been obtained that the action of coldconcentrated hydrochloric acid on starch results in rapid degrada-tion, through the usual successive stages, to maltose, the speed ofthis reaction being much superior t o that of the conversion ofmaltose into glucose. It is shown in the paper now under con-sideration92 that the hydrolysis of starch by acids proceeds 011essentially similar lines to the action of taka-diastase on the poly-saccharide, and that in each of these reactions the glucose finallyproduced is entirely derived from maltose. In this connexion,reference may be made t o an interesting account of the action oftaka-diastase on starch, in which the successive formation ofpa-maltose, aa-maltose, and glucose is clearly illustrated by graphici11ethods.93 Another result, which will not be considered altogethersurprising, has been arrived a t in the course of these investigations.It is now shown that the action of concentrated hydrochloric acidin effecting the auto-condensation of glucose extends to solutionscontaining as little as 1 per cent.of the sugar. This observationweakens the conclusions drawn by Willstatter and Zechmeister 9489 C . S. Hudson and €1. S. Piline, J. Amer. Chem. Soc., 1914, 36, 1571 ; A , ,i, 1148.C. S. HII~SOII, ibid., 1566 ; A., i, 1147.y i A. Harden and W. J. Yoling, Eiochem. J., 1913, 7, 630 ; A . , i, 237.'J2 A. J. Daish, T., 1914, 105, 2053, 2065.y3 W.A. Davis, J. SOC. Dyers, 1914, 30, 249. Ann. Eeport, 1913, 87ORQANIC CHEMISTRY. 93as to the glucose content of cellulose, but, fortunately, does notseem to affect adversely the attempt t o deduce the structure ofcellulose by methylation, which was described last year. From thefact that a definite crystalline trimethyl glucose lias now beenobtained by hydrolysis of methylated cellulose by nieans of highlyconcentrated hydrochloric acid, there seems every reason to hopethat the linkage of the hexose residues in the polysaccharide willsoon be ascertained.95To return to the subject of the diastatic hydrolysis of starch, i tmust be admitted that the process is, in all probability, much morecomplex than is generally supposed. There is even considerabledoubt in the minds of all who have worked with maltose as towhether a perfectly pure specimen of this sugar has ever beenobtained, and the opinion is gaining ground that the substanceusually described as maltose is a mixture of closely related com-pounds.To the evidence already available may be added theobservation that, during the action of malt diastase on starchg'ranules,QG a dextrin is produced the molecular weight of which isnearly equal to that of maltose. Again, many recent experi-ments97 go to show that average specimens of starch consist ofmixtures of compounds which are practically identical in theirproperties, and this may serve t o explain many of the discordantresults to be found in the literature, both with regard to starchand to maltose.Even granting that maltose is a definite chemical individual, itis doubtful if the structure a t present assigned to the compound iscorrect.Direct experimental methods bearing on this question aredifficult to obtain, but reference may be made to a novel attemptto ascertain the structure of maltose, which depends on an applica-tion of Nef's work in the sugar group.98 Curiously enough, theoxidation of maltose by means of alkaline hydrogen peroxide pro-ceeds, to a large extent, without disruption of the disaccharideinto glucose. The essential product thus obtained was a glucosido-glycollic acid, which was subsequently hydrolysed into glucose andglycollic acid. So far as these results go, it would appear thai,Fischer's formula for maltose, if not confirmed, is a t leastsupported.95 W.S. Denham and Miss H. Woodhouse, T., 1914, 105, 2357.97 C. Tanrct, Compt. rend., 1914, 158, 1653 ; A,, i, 665 ; ibid., 1914, 159, 530 ;98 W. L. Lewis and S. A. Buckborough, J. Awaer. C'hesn. Soc., 1914, 36, 2385 ;J. L. Baker and H. F. E. IIulton, iijid., 1529.A., i, 1167A , i, 119994 ANNUAL REPORTS ON THE PROGRESS OF CHEMISTRY.Nitrogen Compounds.Excluding results which are more appropriately discussed undercyclic compounds, or those which bear directly on the problems ofphysiological chemistry, comparatively few definite advances havebeen made in the study of aliphatic nitrogen compounds duringthe past year. Obvious causes have contributed to this fact, but,a t the same time, it must be admitted that a certain amount ofirregularity and the absence of a definite object characterises muchof the synthetical research in this field.Again, as already in-dicated, the extensive and important kork now being carried outin the purine group hvolves reactions which are characteristic ofring structures rather than of open-chain compounds, so that a com-paratively small number of topics are available, on the presentoccasion, for this section of the Report.Carbamides.-Considering the present position of our views re-garding tfie mechanism of processes operative in solution, it is notsurprising to find that the study of the formation and decomposi-tion of carbamide has again been revived.I n this connexion, it may be noted that the earlier work ofWalker and pupils on the change ammonium cyanate + carbamidehas now been completed, so far as the determination of the effectof alcohol on the velocity of the reaction is concerned.Hitherto,on the basis of conductivity measurements, the change has beenregarded as dependent on the interaction of NH,' and CNO' ions,even when 90 per cent. alcohol is used as the solvent, and it isnow shown that for alcohol concentrations increasing from 90 to99.9 per cent. the steady acceleration of the translormationpreviously observed is maintained. This acceleration holdswhether the reaction is referred to the ions of ammonium cyanateo r to the non-ionised salt.99 On the other hand, the addition ofalcohol to aqueous or acid solutions of carbamide exerts a markeddiminution in the velocity of the decomposition into ammonia andcarbon dioxide.1 This result, although apparently unexpected,certainly supports the idea that the change in question involvsthe preliminary formation of ammonium cyanate, and is not adimple hydrolysis.It should be mentioned that the effect of methyl sulphate as amethylating agent has recently been applied successfully tocarbamide and its homologues, with the result that methyl ethersof the corresponding isocarbamides are formed.2 Quite apartfrom the interest attached to this direct formation, from carb-G.J. Burrows and C. E. Fawsitt, ibid., 609.E. A. Werner, ibid., 923.99 J. D. M. Ross, T'., 1914, 105, 690ORGANIC CHEMISTRY. 95amides, of derivatives possessing the iso-structure, the reactions ofthe ethers formed are important from the structural point of view.Thus, when strongly heated, these compounds undergo molecularrupture, the nature of which may be seen from the followingrepresentation of the decomposition of isocarbamide methyl etherin the form of its methyl hydrogen sulphate:H ':..NH2*CH,*HS0, Me*NH,,CH,*HSO, + HN:CO1 /;.ir'%JH,,CH,*HSO, + [Me*OCN + Me-NCO]This dois not exhaust the list of decomposition products, butwill indicate the general grounds upon which the idea is basedthat the constitution of monoalkylisocarbamidm is best expressedby the cyclic structure RN:C<yH3, which, in turn, is a reason- 0Compared with able modification of the formula RN: C<carbamide itself, the tendency of tliiocarbamide to react in simplechanges in accordance with the iso-structure is much more pro-nounced.Several examples may be quoted to illustrate this dis-tinction, and mention may be made of one o r two very emphaticcases which have recently come t o light. By tlhe action of chloralhydrate on carbamide,3 direct aldol condensation results, whichmay involve one or both of the amino-groups, giving the productsN:C/ I ',t\ d M eNIT,OH 'CCl,*CH(OH)*NH*CO*NH, andCCl,*CH(OH)*NR*CO*NH *CH(OH)*CCl,.I n each case the properties of the products are in agreement withthe idea that they are derivatives of normal carbamide, althoughtheir behaviour towards acetic anhydride is somewhat irregular.4On the other hand, thiocarbamide reacts with chloral hydrate inan entirely different manner, as hydrogen chloride is eliminatedduring the reaction, with the ultimate formation ofThe general nature of this reaction is, therefore, similar to thatwhich ensues when monochloroacetic acid acts on carbamides, asin this case, also, the elements of hydrogen chloride are removed,as shown in the following typical example: 5NH,*C(:NH)*S*CCl,*CH(OH),.NHZ*C(:NH)*SH + C?H2Cl.C0,H=NH,*C(:NH)*S*CH,*CO,H + HCl.It would appear, however, from a scrutiny of Feist's work thatF.Feist, F. Nissen, and G. Stadler, Ber., 1914, 47, 1173 ; A., i, 666.' N. G. S. Copyin and A. W. Titherley, T., 1914, 105, 32.' P. C. Ray and F. V. Fernandes, T., 1914, 105, 215996 ANNUAL REPORTS ON THE PROGRESS OF CHEMISTRY.reactions of the latter type proceed only under the mildest possibleconditions, and the complexity of the reactions of thiocarbamidesis further emphasised by the behaviour of these compounds towardshypobromites,G as in certain cases only one atoin of iiitrogen isevolved in the reaction, whilst in others both nitrogen atoms arestable.Again, the action of substit'uted malonic esters on carb-amide and thiocarbamide is utterly distinct,7 although it is to benoted that thiocarbamide reacts with ethyl aminomalonate accord-ing to the normal structure t o give 2-thiouramil,A mino-acids.-A somewhat curious position has arisen in con-nexion with the chemistry of amino-acids, as the synthesis anddecomposition of optically active representatives of this class havepractically ceased.As pointed out last year, this is probablylargely due to the uncertainty with which all synthetical work ofthis nature is invested, owing t o the occurrence of optical inversion,Work in this field is therefore, in the meantime, largely limited t othe identification of naturally occurring amino-acids, the synthesisof simple types, and a study of the general reactions of the com-pounds without reference to their optical behaviour.The optical activity of amino-acids is doubtless a complex factor,owing to the tendency of these compounds to form internal salts,and even the rotation of their metallic salts is affected by the factthat they are hydrolysed in aqueous solution.9 This is shown bythe alterations in rotatory power observed when increasing amountsof acid or of alkali are added t o solutions of an active amino-acid.I n the case of a inonobasic acid, the specific rotation undergoessteady alteration until slightly more than the equivalent amountof acid or alkali has been added, after which, on further addition,the optical value remains constant.With dibasic acids, thesealterations in rotation are even more instructive, and in such cases,also, constant values are only obtained when the amount of addedalkali or acid is slightly in excess of that required for completeneutralisation. I n the paper now referred to, the results are ex-tended t~ a determination of the degree of hydrolysis of salts ofamino-acids, and to a calculation of the basic and acidic ionisationconst'ants of the acids examined by this polariinetric method.The results just described do not seem to be affected by the forma-tion in solution of betaine-like salts, but fresh evidence bearingon this point has been obtained in another investigation, whichti V.von Cordier, Honntsh., 1914, 35, 9 ; A . , i, 258.7 T. B. Johnson and A. J. Hill, J. Amer. Chem. Xoc., 1914, 36, 364 ; A . , i, 330.8 T. B. Johnson and B. H. Nicolet, ibid., 355 ; A . , i, 328.J. K. Wood, T., 1914, 105, 1988ORGANIC CHEMISTRY’. 97deals with the action of diazomethane on amino-acids and theiracetyl derivatives.1° It appears that the reagent is without actionon the free acids, so that, under the conditions of the experiment,a functional carboxyl group is absent.On the other hand, methyl-ation proceeds normally in the case of substituted amino-acids,indicating the presence in these compounds of a normal carboxylgroup. The conditions under which a-amino-acids react as cyclicstructures are, so far, imperfectly understood, and i t is desirablethat further research, involving physical methods, should bebrought to bear on the subject, as i t is no doubt intimately con-nected with the mechanism of the Walden inversion as experiencedwith these compounds. The action of nitrous acid on a-amino-acids is, of course, comparatively regular, but is occasionally accoin-panied by inversion or racemisation, and Fischer has recentlyencountered another example which emphasises this point.11Obviously, the elimination of the amino-group will follow adifferent course, according as the amino-acid is present in either ofthe two structural forms shown below:and in optical studies of this change it would be highly desirableto be in possession of an accurate physical method of determiningthe relative amounts of these isomerides present during the de-composition.It is conceivable that in this way some furthergeneralisations underlying this important change may yet berevealed .The above structural views may also serve to, explain some ofthe discordant reactions of anhydrides, derived from amino-acids,which have recently been described. For example, the observa-tion that aminoacetaldehyde is produced by the electrolytic reduc-tion of glycine anhydride12 is not consistent with the normalformulaNH-COas the group *CO*NH* is converted in such reactions into*CH,*NH,. It has, in fact, been suggested that diketopiperazineis more accurately represented by the formulaCH,<CO.NH>CH2,NH*NHa structure which would easily be derived from the cyclic formof glycine.JAMES COLQUHOUN IRVINE.lo A. Geake and M. Nierenstein, Zeilsch. physiol. Chin., 1914, 92, 149 ; A., i,1057. l1 E. Fischer and R . Ton Gravenitz, A?znaZen, 1914, 406, 1 ; d., i, 1057.l2 G. W. Heimrod, Ber., 1914, 47, 338 ; -4., i, 327.CH,<~*--CO>CH27REP.-VOL. X I . 98 ANNUAL REPORTS ON THE PROGRESS OF CHEMISTRY.PART II.-HOMOCYCLIC DIVISION.IN spite of the decrease in the output of literature during the lastthree months of the period under review, the earlier portion ofthe year showed no falling off in the number of papers dealing withthis branch of Organic Chemistry.It will be observed that nowriter of the reports on this division of the Science, since theirinception, has failed t o remark on the difficulties of the task ofselection with which he was confronted, and has offered apologiesfor his many omissions. Indeed, no report could do justice to themass of material which did not reproduce, in their entirety, theabstracts published in the Journal. To compress into1 a few pagesa review of these numerous researches, when none of them can bedescribed as brilliant and most of them are good, necessitates aselection which, in the nature of things, must be largely haphazard.The writer has therefore selected those communications whichseemed to him t'o be the more important, but he is fully consciousof the fact that he has neglected much that is of value.One of the most striking features of the literature under reviewis the large proportion which is devoted to the mere statement offact, and the small amount which deals with the principles under-lying these facts.This condition is, of course, a natural outcomeof the great increase in the quantity of literature published duringpast years, which, becoming an intolerable burden on the financialresources of the publishing societies, has necessita.ted the shorten-ing of papers, sometimes t o such a degree as to render them hardlyintelligible. I n any case, the discussion of the results of otherworkers, unless concrete facts are disputed, is either reduced to aminimum o r entirely ignored. This is particularly noticeable ininternational work, and one could easily believe that, in certaincountries, no notice whatever is taken of papers published abroadwere it not for the fact that internal evidence to the contrary isoften forthcoming.Another disquieting feature is the increasing number of state-ments which, on subsequent investigation, have proved t o be wrong.Many of these, no doubt, are due to clerical errors, but no onewho is conversant with the literature of modern organic chemistrycan have failed to meet with cases which can only be ascribed t oother causes.I n the ensuing report, the various data have been collected, sofar as possible, under general headingsORGANIC CHEMISTRY.99Nitration.I n view of the importance attaching to trinitrotoluene (T.N.T.)a t the present time, a communication by W. Will1 is of interest.Of the six possible trinitro-derivatives of toluene , only three areknown, namely , the pure commercial article, the so-called u-trinitro-toluene (m. p. 80*6O), which is 2 : 4 : 6-trinitrotoluene; y-trinitro-toluene (m. p. 104O), obtained from m-nitrotoluene, which is2 : 4 : 5-trinitrotoluene ; and the P-isomeride (m. p. 112O) , formedby the nitration of both 2 : 3-dinitrotoluene and 3 : 4-dinitrotoluene,which is 2 : 3 : 4-trinitro€oluene. The three isomerides are prac-tically of the same value as explosives.A search for higher nitrated derivatives proved unsuccessful,because if the reaction is promoted by heat or pressure, eithertrinitrobenzoic acid, or even tetranitromethane, is obtained.It ispointed out that the intense colour of the latter substance is some-times noticed in the factory, but that trinitrobenzoic acid, owingt o the solubility of its salts, has hitherto escaped detection; itspresence is likely t o be a source of danger. The author has ex-tended his researches to benzene, and considers t'hat it is extremelydoubtful whether a higher nitrated derivative than trinitrobenzeneexists.An interesting example of the failure of the aldehyde group t oexert its usual directive influence into the meta-position isdescribed by W.H. Perlun and R. Robinson2 in the nitration ofo-veratraldehyde. The formula of the aldehyde (I) shows thepositions a and b to be symmetrically placed as regards themethoxy-groups :Me0 Me0 Me0XeO/)CIIO M~O/)CHO Mo0f)GHO'\P \/ \?*2(111.)b NO2(1.1 (11.1It might reasonably be expected, therefore, that the m-nitro-derivative (11) would be formed on nitration, whereas the soleproduct is the o-nitro-derivative (111).A method for preparing nitro-compounds from amino-compoundsby the spontaneous decomposition of the nitrites of the correspond-ing diazo-compounds is described by W. Korner and A. Contardi.3The process appears to be of some value, since, for example,2 : 6-dibromosulphanilic acid can be quantitatively converted byBer., 1914, 47, 704; A., i, 509.Atli R.Accad. Limei, 1913, [v], 22, ii, 625 ; d . , i, 263.T., 1914, 105, 2376.H 100 ANNUAL REPORTS ON THE PROGRESS OF CHEMISTRY.this means into’ 2 : 6-dibromonitrobenzeiie-4-sulphonic acid, and twosubstituted dinitrobenzenes can also be prepared from the corre-sponding nitro aniliiies.The bromine-titration method, applied with success to ethyl aceto-acetate and similar substances, has been extended to an investiga-tion of the desmotropism exhibited by nitro-compounds and nitro-ketones.4 There appears to be a fundamental difference betweenthe phenomenon as shown by this type and that exhibited by theketo-enol compounds. Thus, whereas the enolic form of, forexample, ethyl acetoacetate is favoured least by water, more bymethyl alcohol, and still more by ethyl alcohol, the reverse is thecase with the isonitro-compounds. This observation is in accordancewith the rule that the proportions of the two isomerides dependon their solubilities in the solvent, for enols are less soluble thanketones in water, but aci-compounds are more soluble than truenitro-compounds.A z o-compouids.K.H. Meyer and S. Lenhardt5 have already proved that, con-trary to the usually accepted idea, not only phenols, but phenolicethers, couple with diazonium salts to form azo-compounds.Investigation now shows that, spe’aking generally, the introduc-tion of negative groups into the diazonium salts increases theiractivity towards coupling with phenolic ethers.Thus thediazonium salt derived from 2 : 4-dinitroaniline couples readily withanisole and with phenetole, and combines immediately with theethers of resorcinol and a-naphthol. On the other hand, the intro-duction of negative groups into the phenolic ethers diminishes theirtendency to couple, whilst positive radicles, especially alkyloxy-and alkyl groups in the meta-position, increase the tendency toform azo-compounds.K. v. Auwers and F. Michaelis’ confirm the latter statement,but point out that a free para-position is essential, and that thespeed of the reaction is increased when there is a second alkylgroup in the ortho-position, and is a t its maximum when the twoalkyl groups are both meta to the ether group.There is a great deal to be said for the view advanced by K.H.Meyer6 and his co-workers to the effect tllat this coupling repre-sents an addition of the diazo-salt to a double bond of the secondcomponent. They point out that the methyl ether of 9-methyl-anthranol, which possesses no “active” hydrogen atom in theK. €1. Meyer anil P. Wertheimer, Ber., 1914, 47, 2371 ; A!., i, 1061.Annnlen, 1913, 398, 66 ; A., 1913, i, 723.K. H. Meyer, A. Ir>chick, and H. Schliisser, Ber., 1914, 47, 1741 ; A . , i, 882.Bey., 1914, 47, 1275; A , , i, 744ORGANIC CHEMISTRY. 101ortho- or para-positions, can nevertheless couple with p-nitro-benzenediazonium hydroxide, in accordance with the scheme :They therefore regard the coupling of a diazoniuin salt. with apheiiolic ether as taking place in the following manner:H \ / OMe\/’ -= \/ -+NR:N/\ -/\OH / \/ \for the fact that, in some cases, the This exDlanation also accounts Icoupling of the diazo-compound and the alkyloxy-compound yieldsa large proportion of the free hydroxyazo-derivative.The supposed isomeride of azobenzene obtained by C.V. andR. A. Gortner? which melted a t 25O, and was found t o be con-vertible into ordinary azobenzene (m. p. 68O) by means of dilutehydrochloric acid, has been investigated by H. B. Ilartley andJ. M. Stuart,g and is found to be a solid solution of azobenzenein azoxybenzene.It is well known that phenol and o- and ?n-cresol form bisazo-compounds in concentrated, but not in dilute, solutions ; thymoland carvacrol also react in the same way.It is now pointed out7that the tendency to the formation of bisazo-compounds rises withthe number of alkyl groups in the phenol molecule, whereas thepresence of other substituents, such as the nitro-, carboxyl, andcarboxylic ester groups, destroys, not merely the tendency to, butfrequently the possibility of, such condensation. It is also statedthat, as a general rule, negative substituents diminish the re-activity of the phenol and that the formation of the bisazo-com-pound is contlrolled not so much by the speed of coupling of thephenol as of the monoazo-compound. Bisazo-compounds are bestformed in alkali hydroxide solutions, not so well in the presenceof alkali carbonates, and scarcely a t all in acetic acid.J.Anzer. Chcm. Soc., 1910, 32, 1294 ; A, 1910, i, 790.3:, 1914, 105, 309102ThetionalANNUAL REPORTS ON THE PROGRESS OF CHEMISTRY.so-called diazophenols can be representesd by three constita-formulz,R<ONiN(1.1 (11.) (111.)Wolff’s researches have rendered formula I improbable, and he,with Hantzsch, favours the third constitutional formula. A.Klemenc10 now brings forward evidence in favour of formula 11.IIe argues that whilst the explosive character of these compoundsagrees with either constitution, the fact that the substances dissolvein concentrated acids and can be recovered on dilution wouldsuggest that they cannot be quinone-diazides of formula 111. Itseems highly improbable that any definite chemical evidence willever be obtained which will differentiate between compounds offormulae I1 and 111, although it is to’ be remembered that, as theBlomstrand diazonium formula is a t present fashionable, i t appearsreasonable to assign formula I1 t o the diazophenols, since they havemany characteristics in common with the diazonium salts.Colour and Constitution.As the result of the comparison of a large number of colouringmatters, a theory is advanced,ll by which it is stated that thosedyes which are quinonoid in all possible tautomeric forms exhibita deep colour, whereas those which can be represented by formulznot containing the quinonoid structure are paler in shade.Although this working hypothesis is amply supported by the factsadvanced, yet i t is evident that the generalisation fails t o accountf o r many changes in the intensity of colour which are met within compounds of closely similar structure and molecular weight.One instance of this kind may be mentioned, namely, the occur-rence of blue colours by the coupling of naphthol derivatives withdiamino-bases which yield substantive cotton dyes, whilst the samesecond component yields red colouring matters with those diamino-bases, of closely related structure, which do not produce substantivecotton colours.The suggestion originally made by J.T. Hewitt and A. D.Mitchell 12 that substances of the t*ype of pnitrobenzeneazo-a-naphthol and its potassium salt have structures represented by theformuke/-\ /-\\-/ \-/\-/ \-/ \-/ \-/ O,N,’-\N:N*/ \OH and KO,N:/ -\:N*N:/ \:Ol o Bcr., 1C14 47, 1407 ; A , , i, 743.l2 T., 1906, 89, 19 ; ibid., 1907, 91, 1251.l1 E.R. Watsofi, T., 1914, 105, 759ORGANIC CHEMISTRY. 103has received support by an extension of the work13 to pamino-acetophenone and paminobenzophenone in combinatJon withphenol, 0-cresol, and a- and 6-naphthol. The results are in accord-ance with the view that these compounds and their salts shouldbe represented by type A, whereas the corresponding phenyl-hydrazones are t o be expressed by type B:j' CH,*CO*C,EI,*N:N* @,H4* OH.1 CH,= c (OK) : C,H4: N-N: C,H4: 0.A.CH,*C( :N*NH*C',~,)*C,H,*hT:N~C,H,~OH.CH,*C( :N*NH*C6H,)*C,H,*N:N=C6H4*OK.The large amount of work on the nitroaminophenols which hasbeen carried out by R.Meldola and his collaborators f o r manyyears past has led to the important conclusion14 that the conditionessential to the production of deep colour is the ortho-position ofa nitro-group with respect to an amino- or substituted amino-group.The presence of additional nitro-groups in the nucleus has gener-ally the effect of increasing the intensity of the colour. It isevident that the appearance of colour in these cases cannot bedue to the mere presence of nitro- and amino-groups in the nucleus,but must be caused by some more profound change in structure.The compounds named are theref ore given the structures indicatedby the numbered formulze : 3 : 5-dinitro-paminophenol (I), 2 : 3 : 6-trinitro-p-aminophenol (11), and 2 : 3 : 5-trinitro-panisidine (111) :OH OH 0-CH,(111.)(Red.)An important illustration of the property of colour in relationto tautomerism is brought forward by R. Meldola and W. F.Hollely15 in connexion with the compound formed by the inter-action of 2 : 3 : 5-trinitroaceto-panisidide and aniline. The freecompound can be obtained in pale, ochreous needles (IV) or deepred scales (V), the two forms being interconvertible. Both modifi-cations dissolve in alkali, forming a deep orange solution of com-pound V I :It J. '1'. Hewitt, Miss G. R. Mann, and F. G . P o p , T., 1014, 105, 2193.l4 R. Meldola and W. F. Hollelg, ibid., 414.l5 Bid., 977104 ANNUAL REPORTS ON THE PROGRESS OF CHEMISTRY.Friedel and Crafts’ Reaction.I n the last volume of these Reports10 attention was drawn to thelarge amount of work which has been published recently on thevarious applications of this reaction, and a t the same time a valu-able summary was given respecting our knowledge of the mechanismof the process.During the period under review, several papershave appeared which have an important bearing on this question.S. C. J. Olivierl’ has conducted certain dynamic researches on thecourse of the reaction, p-bromobenzenesulphonyl chloride beingused, since it not only reacts with measurable velocity, but is notdecomposed by water at the ordinary temperature. Evidence wasobtained of the formation of the additive product showing that,partly a t any rate, the reaction proceeds in accordance with thescheme :C,H,Br~SO,Cl,AlCl, + C,H, -+The following conclusions were drawn :(1) The acid chloride reacts solely in the form of the compoundC,H,Br*SO,CI,AlCl,.(2) One molecule of aluminium chloride cannot transform morethan one molecule of acid chloride.(3) The reaction is unimolecular with respect to the compoundC,H,Br SO,Cl, A1 Cl,.(4) The reaction constant (when an excess of aluminiumchloride is not used, and if K is calculated for the compoundC,H,Br*SO,Cl,AlCl,) is proportional to the concentration of thealuminium chloride.(5) The constant is greatly increased by an excess of aluminiumchloride.These facts are most readily explained on the assumption thatthe acid chloride is activated proportionally t o the concentrationof the combined aluminium chloride. When benzene is replaced byits derivatives, reaction occurs with diminishing rapidity in theorder benzene, bromobenzene, chlorobenzene.No reaction wasobserved with nitrobenzene. It should be added that the aboveC,H,Br*SO,*AlCl, + C,H,CI + HCI.?6 Ann. Xcport, 1913, 97. l7 Rec. tmv. cJ&n., 1914, 33, 91 &A,, i, 818ORGANIC CHEMISTRY. 105conclusions receive support from the observation of C. R. Rubidgeand N. C. Qua,18 who find that, in the reaction between benzoylchloride and benzene in the presence of aluminium chloride, theyield is considerably diminished if the amount of aluminiumchloride is reduced.On the other hand, B. N. Menschutkin19 has continued his ex-periments on the influence of antimony trichloride and antimonytribromide on condensations of the Friedel and Craft type.Thefirst phase in the reaction invoIves the formation of a compoundof the antimony salt with the hydrocarbon, 2SbC1,,C6H,R. Thiscompound is then acted on by, for example, benzoyl chloride,yielding a compound of antimony chloride with the ketone,SbCl,,C,H,*CO-C,H,R, which decomposes into its constituent8 atthe temperature of the experiment. The antimony trichloride thusliberated may then react with fresh quantities of hydrocarbon andbenzoyl chloride.ArH + C,H,*COCl + SbC1, = Ar-CO*C,H, + HC1+ SbCl,,proceeds t o an end, and gives results in accordance with a bimole-cular reaction, the hydrocarbon and benzoyl chloride being takenalways in molecular proportions. The velocity of the reactionvaries directly as the square of the concentration of the antimonytrichloride.It is certainly difficult to bring the results of these two investi-gations into line, and it is evident that the problem of themechanism of this reaction is still unsolved.It should be addedthat J. Boeseken and M. C. Bastetzo bring forward furtherevidence in supportl of their view that the Friedel and Craft re-action takes place when three molecules are present, the first ofwhich is unsaturated, the second of which can be activated to suchan extent that i t is decomposed during the reaction into two parts,which then unite with the first molecule, and the third of whichcan activate the two molecules. The authors describe a series ofcondensations between benzene and various chloro-derivatives ofethylene in which, in the first stage, the molecule of benzene isdisrupted, and then forms a compound with the unsaturated mole-cule; the latter molecule then becomes disrupted in its turn, andcombines with a second or third molecule of benzene, which nowbehaves as if it were unsaturated.It has been shown21 also that aluminium chloride may be usedas a means of effecting the' elimination of water in certain typesl8 J.Anzer. Chena. Soc., 1914, 36, 732 ; A . , i, 589.lY J. Iiuss. Phys. Cham. SOC., 1913, 45, 1710; 1914, 46, 259 ; A . , i, 188, 673.2o Bec. trnv. chim., 1913, 32, 184 ; A., i, 156.21 G. B. Frankforter arid W. Kritchevsky, J. Amer. Chem. SOC., 1914, 36, 1511 ;The tot21 reaction, expressed by the equationA,, i, 1059106 ANNUAL REPORTS ON THE PROGRESS OF CHEMISTRY.of condensation.Thus hydrocarbons of the aliphatic, benzene,naphthalene, and anthracene series have been condensed, not onlywit,h chloral, but also with chloral hydrate and bromal, thereactions, which are carried out a t Oo, being represented by theequation2RH + CC13*CH0 = CHR,*CCl, + H,O.Biphenyl Series.The stereochemical formula for diphenyl originally proposed byKaufler22 to account for the production of cyclic compounds ofthe type of phthalylbenzidine,Q,H,*NH*CO*C,,H,*N H* CO >C,H,has apparently received support by recent investigations in thediphenyl series.23 J. C. Cain, A. Coulthard, and Miss F. M. G.Micklethwait 24 have shown already that two distinct o-dinitro-benzidines exist, and the cause of this isomerism becomes apparentif it is assumed that in the Kaufler formula,2’- 3’there is no free rotation between the two benzene rings.The twodinitro-derivatives are therefore 3 : 3’- and 3 : 5/-dinitrodiphenyl.An extension of the work to o-tolidhe25 has led t o the isolationof two, dinitro-o-tolidines, and in this instance it was found possibleto convert one of them into the other.Considerable interest a.ttaches to the fact that the four possible*g2 Annalen, 1907, 351, 151 ; Bcr., 1907, 40, 3250 ; A., 1907, i, 794.2 3 J. C. Cain and Miss F. M. G. Micklethwait, T., 1914, 105, 1438.24 T., 1912. 101, 2298.2.5 J. C. Cain aiid hliss F. M. G. Micklethwait, ibid., 1914, 105, 1442.* The writer is under the impression that there are six possible isomerides and thatthe two formu18shotild be included.rings and the argument is not affected by their exclusion.These compounds contain, however, two dissimilar benzenORGANIC CHEMISTRY.107m-dinitro-derivatives of o-tolidine required by the Kaufler formulahave been isolated, namely,and it is csrtainly remarkable that no less than three of thesegive distinct carbazoles on reduction. This involves, in a t leastone case, the production of tram-linking, thus :Such linking is, of course, known, since the trans-forms of manydialkylsuccinic acids, compounds in which the ring-f ormingelements are separated in the same manner as in these reducednitro-compounds, yield their own anhydrides, but if this is thecase the trans-carbazole should be readily convertible into its cis-isomeride, which must be identical with one of those derived fromthe other two bases.It is to be noticed, moreover, that Cain andMiss Micklethwait 26 prepared condensation products of benzidinewith o-diketones to which they assign the formula:/-'\N: CRan assumption which, if correct, would give strong support to theKaufler formula. J. Kenner and Miss A. M, Mathews27 point out,however, that in no case was it found possible t o free these con-densation products from the two molecules of alcohol or othersolvent which always accompany them, and they remark thatTauber28 does not appear t o have experienced any difficulty inobtaining similar free condensation products from 2 : 2I-diaminodi-26 LOC.cit., p. 1440.28 Ber., 1892, 25, 3287 ; 1893, 26, 1703 ; A . , 1893, i, 96, 588.T., 1914, 105, 2473’108 tlNNUAL REPORTS ON THE PROGRESS OF CHEMISTRY.phenyl. They also argue 29 that i f Cain and Miss Miclrletliwait’sview as to the isomerism of the’ two dinitrobenzidines is correct, it isto be anticipated that the prevention of rotation of the two benzenenuclei would also be exhibited in 2 : 2’-diphenyldiacetic acid. Suchis, however, not the case, since whilst diethyl diphenyldiacetatefurnishes ethyl dibenzocycZoheptadienecarboxylate,”O diphenyldi-acetyl chloride can be converted into hydroxypyrene (11) 31 :/\I bH(C0, EX)--\y o ‘4IThis argument is, of course, by no means conclusive, since i t isconceivable that the reagents used niight have caused rotation ;moreover, i t must be remembered that’ Cain and Miss Mickle-thwait 32 prepared a condensation product from tolidine and glyoxalwhich, although not crystalline, gave a nitrogen content correspond-ing with that required by the free condensation product.Furtherwork on this subject will be awaited with interest, because i t isapparent that more evidence is desirable before the Kauflerformula can be accepted.Triairylm e thyls.An admirable address on the triarylmethyls has appeared,33 inwhich the arguments for and against the conception of freeradicles are discussed, and in this connexion certain experimentsdescribed by W. Schlenk and E. Marcus34 are of importance.These investigators find that metals such as sodium or potassiumwill combine with the free organic radicles if the reaction is carriedout in ether solution in an atmosphere of nitrogen.Of the triaryl-methyls, only triphenylmethyl presented difficulties, which were,however, overcome, and sodium triphenylmethyl, CPh,Na, wasobtained as a brick-red mass very sensitive t o air. The structureof this substance follows from the formation of triphenylmethaneby the action of water or hydrochloric acid, and of triphenylaceticacid by the action of carbon dioxide. Methyl iodide and benzylLOG. tit., p. 2474.31 R. Reitzenbock, Alomtuk., 1913, 34, 199 ; A., 1913, i, 259.32 LOG. cit., p. 1441.33 M. Goniberg, J. Amcr. Chem. SOL, 1914, 36, 1144 ; A., i, 823.y4 Ber., 1914, 47, 1664 ; A,, i, 843.:j0 J.Kenner, T., 1913, 103, 615ORGANIC CHEMISTRY. 109chloride react at once, forming triphenylethane and as-tetraphenyl-ethane respectively.T?he Cinrmmic Acids.The explanation advanced by E. Biilmannss t o account f o r theoccurrence of three modifications of cis-cinnamic acid, namely, thatthe acid is trimorphous and can be isolated as allocinnamic acid(m. p. 68O), isocinnamic acid (m. p. 58O), and isocinnamic acid(m. p. 42O), is at the present time very widely held. H. Stobbeand C. Schonburg36 have now conducted a large number of experi-ments on the structure of these three acids, and have corm to theconclusion that they are, in reality, chemical isomerides. Thegeneral impression produced by the paper is that the experimentsdescribed can, for the most part, be explained on the Biilmanrhypothesis. It would appear desirable, before reopening this ques-tion, to frame some really adequate theory to account for morethan two stereoisomeric forms of cinnamic acid.Condehsa tion.Many papers have appeared during the period under review inwhich the formation of various types of condensation products isdescribed, but they deal, for the most part, merely with applica-tions of old methods to new conditions.Some, nevertheless, marka distinct step in advance; thus it has been found37 that manyacid chlorides condense with diphenylketen a t the ethylenic link-ing, forming condensation products, of which diphenylmalonylchloride, COCl-CPh,*COCl, formed from diphenylketen and oxalylchloride, may be taken as a type.The reaction does not appear,however, to have a very extended application, since only few acidchlorides react without yielding secondary products of indefinitestructure.The large number of compounds which are a t present knownhaving one or other of the alkali metals directly combined withcarbon, and which are therefore of considerable importance insynthetic work, has besen increased by the extension of the workof W. Schlenk and his collaborators on the formation of themetallic ketyls38 t'o a study of the conditions favouring the additionof the alkali metals to compounds containing the complexes C:C,C:N, and NzN.39 The reactions are carried out in ether solution,xj Ber., 1909, 42, 182, 1443; 1910, 43, 568 ; A ., 1909, i, 155, 382 ; 1910, i, 346.36 Annalev. 1913, 402, 187 ; A , , i, 173.57 H. Staudinger, 0. Gdliring, aid M. R. Schiiller, Rer., 1914, 47, 40 ; A., i, 285.38 W. Schlenk and A. Thal, A., 1913, i, 1205.3g W. Schlenk, J. Appenrodt, A. Michael, and A. Thnl, Ber., 1914, 47, 473 ;A , , i 396110 ANNUAL REPORTS ON THE PROGRESS OF CHEMISTRY.and appear to have a general application; thus, the C:C compound,stilbene, forms the disodium derivative CHPhNa-CHPhNa, thestructure of which is proved by the production of s-diphenylethaneby the action of water, and of the sodium salt of s-diphenylsuccinicacid when it is treated with carbon dioxide. The C:N compound,benzylideneaniline, yields the disodium salt,NPhNa*CHPh=CHPh*NPhNa,which with carbon dioxide gives the sodium salt of the acid,CO,H-NPh*CHPh*CHPh*NPh*CO,H.As a type of N:N compound, azobenzene was chosen, and in thiscase an equimolecular mixture of azobenzene and the dipotassiumderivative, NPhK-NPhK, was formed, from which the potassiumsalt, CO,K*NPh-NPh*CO,K, was produced by the action of carbondioxide.The work initiated by H.Stobbe on the condensation of cyclicketones with ap-unsaturated ketones has been extended 40 to include,the condensation of cyclopentanone and 3-methylcyclohexanonewith the ethyl esters of a-benzylidenebenzoylacetic acid anda-benzylideneacetoacetic acid. I n the case of 3-methylcyclo-hexanone, no condensation could be effected, but cyclopentanonecombines with ethyl a-benzylidenebenzoylacetate in the norinalmanner, yielding the compound I, whereas the condensationbetween the ketone and ethyl a-benzylideneacetoacetate yields ipcompound which is thought to be the bicyclic ketone, 11:An interesting condensation is described by M.Losanitsch,41 inwhich aldehydes and ketones are caused t o combine with saturatedlactones. The condensations are effected in ethereal solution inthe presence of alcohol-free sodium ethoxide, and may be illustratedby the formation of a-benzylidenevalerolactone,from benzaldehyde and valerolactone.The Grig~zurd Reaction.-The question as t o the possibility ofmore than one halogen atom, contained in the molecule of a react-ing substance, of combining with the Grignard reagents, has beeninvestigated by E.Votdek and J. Kohler,42 who find that pdi-iodobenzene yields pdi-iododiphenyl, benzene, and diphenyl, the‘O H. Stobbe, A. Schwyzer, aiid G. S. Crnikshnnks, J. pr. Chem., 1914, [ii], 89,184, 189 ; A., i, 540, 541.No?intsh., 1914, 35, 311 j A . , i, 693.42 Ber., 1914, 47, 1219 ; A., i, 763ORGANIC CHEMISTRY. 111formation of the last-named two substances being due to thedecomposition of ths dimagnesium compound, as shown by thefollowing scheme :I n t rani ol e cu la r Change.Whilst no new examples of the migration of atoms or groupshave been recorded during the present year, several of thosepreviously described have either been supported by further evidenceor have been confirmed by the isolation of definite intermediateproducts.The &figration of para-Halogen A toms in Phenols.-The dis-covery by P.W. Robertson43 that in the nitration of, for example,6-bromothymol (I), the nitro-group displaces the halogen atom,which then migrates to the ortho-position, thus :Br NO2/\'\/ CHI l:f13 /\CH3c 3 4 j 4OH OH(1.)has led to the investigation of other instances of the same kind.It is now found 44 that; whereas 6-bromo-4-nitro-m-cresol (11)behaves normally on nitration, 4 : 6-dibromo-m-cresol (111) yields2-bromo-4 : 6-dinitro-m-cresol (IV) :Br Br/\OH, + N02(/N02OH OHNO"(111.) UV.1The Migration of Acid R esidues.-The phenomenon of migrationof the acetyl group from oxygen to nitrogen during the reductionof certain acetylated azo-compounds has been established by K.v.Auwers and his collaborators. The migration seems $0 be affected,45 T., 1908, 93, 793 ; P. W. Robertson and H. V. A. Briscoe, ibid., 1912, 101,44 I. R. Gibbs and P. W. Robertson, i b i d . , 1914, 105, 1885.1961112 ANNUAL REPORTS ON THE PROGRESS OF CHEMISTRY.or entirely hindered, by the' presence of certain groupings in theazc-molecule, and does not occur with the corresponding benzoylderivatives. I t has been shown4s that the presence of a substituentin either nucleus prevents the transformation except in oneinstance, in which a methyl group occurred in the para-positionof the benzene ruclms. Recent work46 shows that the ethyl groupin the same position also permits the migration.Aromatic from Hydroaromatic Substances.-The remarkabletransformation of a hydroaromatic compound into a true aromaticderivative which is illustrated by the formation of brominatedxylenols from dimethyldihydroresorcin by the action of phosphoruspentabromide,47 has been further investigated by A.W. Crossleyand Miss N. Renouf.48 It is evidently the intermediate compounddibromodimethylcyclohexenone (I), that gives rise t o the aromaticderivatives, since when this substance is treated with potassiumhydroxide ii; is converted into 5-bromo-o-3-xylenol (11) mixed withsmall quantities of 4 : 5-dibromo-o-3-xylenol (111) :C(C*,)2. CH3/\ /\OH,CH3c=*/ ,CH, --,Br!,,)OH t- BrC\/CO "'?/OH/\CH8CBr Br(11.1 (1.) (111. )The Beckmann Rearrangement.-J. Stieglitz and P. N. Leech 49put forward an explanation of the Beckmann change, which differsfrom that of Hantzsch in assuming that.a t on0 stage a compoundis produced containing a nitrogen atom having free residual valen-cies, thus:/H\ClCRRINOH + CRR':N-OH -+ CI1R':NCI: --+ CRC1:NR'The last substance then undergoes hydrolysis in the usuallyaccepted manner with the formation of an amide. This viewis based on the similar behaviour of triarylmethylhydroxyl-amines, for example, triphenylmethylhydroxylamine, CPh,*NH*OH,which on treatment with phosphorus pentachloride in etherealsolution is almost quantitatively converted into benzophenoneanilhydrochloride, CPh,:NPh,HCl.45 Annnlen, 1909, 365, 278 ; A . , 1909, i, 436.46 K. v. Auwers and F. Michaelis, Bcr., 1914, 47, 1297 ; A ., i, 747.47 A. W. Crassley and H. R. Le Sueur, T., 1903, 83, 110.48 T , 1914, 105, 166.49 J. Amer. Chem. Soc., 1914, 36, 2 i 2 ; A . , i, 268ORGANIC CHEMISTRY. 113Cdalyiic Beactions.Ths hydrogenation of diarylacetones and aryl alcohols has beeninvestigate'd by P. Sabatier and MI. Murat.50 Benzophenone and itshomologues readily undergo hydrogenation in the presence, ofslightly active nickel at temperatures from 300° to 350°, givingthe corresponding diaryl-hydrocarbons. A t lower temperatures andwith very active nickel, the aromatic nucleus also undergoes hydro-genation.A. Mailhe51 has prepared a number of new ketones by passingthe mixed vapour of several pairs of acids over ferric oxide a t470-480'. I n this way benzoic and phenylacetic acids yieldphenyl benzyl ketone, CH2Ph*COPh, acetic and anisic acids yieldaiiisyl methyl ketone, and so forth.The method, which can alsobe carried ou't in the presence of manganous oxide,52 seems to beof general application. It is pointed out53 that manganous oxidecan also be used with advantage in place of titanic oxide for thepreparation of aldehydes from the mixed vapours of formic withalipha€ic or arylacetic acids. An application of this cat$alyticmethod of condensation is also seen in the formation of certainether oxides of carvacrol.54 It is found, f o r example, that whenan equimolecular mixture of phenol and carvacrol vapours is passedover thorium oxide a t 470480°, phenyl carvacryl oxide is formed.An interesting observation has been made by E.Knoevenagel,55who finds that in the preparation of thiodiarylamines by the actionof sulphur on diarylamines, the addition of 0-05-2 per cent. ofiodine not only causes the reaction t o take place a t a lower tenipera-ture, but also yields a purer product, and considerably shortensthe time required for the completion of the reaction. A similareffect has been observed in the following reactions: (1) The forma-tion of arylnaphthylamines by the condensation of aromatic amineswith the naphthols and naphthylamines; (2) the alkylat'ion ofaniline and a-naphthylamine by the direct action of alcohols ;(3) anil-formation with ketones and aromatic amines; (4) sulphona-tion; and (5) oxidation.It has been found56 that the bromination of benzene proceedsa t a lower temperature in the presence of manganese, and that theratio of recovered benzene to brominated benzene is 8 : 9 in theLo Compt.rend., 1914, 158, 760 ; A . , i, 548.51 Bull. SOC. chiin, 1914, [iv], 15, 324 ; A . , i, 548.52 P. Sabatier and A. Mailhe, Compt. rend., 1924, 158, 830 ; A . , i, 547.53 Ibid., 985 ; A., i, 547.55 J. p r . Chem., 1914, [ii], 89, 1 ; A , , i, 519.5(i L. Gay, F. Dncclliez, and A. Raynaud, Conzpt. nmd., 1514, 158, 1804 ;54 Ibid., 608 ; A . , i, 403.A., i, 946.REP.-VOL. XI. 114 ANNUAL REPORTS ON THE PROGRESS OF CHEMISTRYpresence of manganese and 10 : 3 in its absence. Similar effectsare observed in the bromihation of toluene and the xylenes in thecold, but a t higher temperatures better results are obtained withoutmanganese.Ketones.The import'ant process for producing 1 : 2-diketohydrindene bythe hydrolysis of oximino-1-hydrindone by means of formaldehydeand hydrochloric acid 57 has now 58 been applied, with even betterresults, to the preparation of 1 : 2-diketo-5 : 6-dimet'hoxyhydrindene(I) and 1 : 2-diketco-5 : 6-methylenedioxyhydrindene (11) :(1.) (11.1These substances may very possibly find an application in solvingthe difficult problem of the synthesis of brazilin and its derivatives.An important addition to the number of desmotropic substancesrelated to the phthaleins has been made by 0.Fischer andE. KOnig,bQ who have succeeded in preparing a fluoran derivative,in both the lactlone and quinonoid forms, by the action of phthalicanhydride on 1 : 6-dihydroxynaphthalene.Since this substance,from its mode of formattion, must be derived from either an US- o ra Ba-naphthafluoran, and since it yields bisazo-dyes which do notdissolve1 in alkali, it follows that the lactone and quinonoid formsare represented by formula: I11 and IV respectively :HO/\ /\:0 Ho{\l 0 1 /\OH I I l O l I\/\/\/\/ \/\/\/\/I I I I\/'\/\/I I I I\/\/\/ c c:/\/ \C6H,*CO*0(111.)The lactone form is unstable, and can only be obtained colourlessin solution or, in solid form containing two molecules of thesolvent (C,8H,60,,2Et,0 ; C,,H,605,2COMe,, and so forth). Whendeprived of the solvent of crystallisation, it passes over into the redquinonoid form.A new method of methylatlion applicable t o aliphatic and hydro-57 W.H. Perkin, jun., W. M. Roberts, and R. Robinson, T., 1912, 101, 232.58 Ibid., 1914, 105, 2405.59 Bcr., 1914, 47, 1076 ; A., i, 712ORGANIC CHEMISTRY. 115aromatic ketones and esters of ketonic acids has been described.60The ketone is converted into the hydroxymethylene compound (Vj,which is t’hen reduced by hydrogen in the presence of colloidalpalladium to the methyl derivative (VI):-c:o --c: 0 -C:O -C:O-C:CH.OH -CH*CH,*OH --O:CH, - I - I-CH*CH, I 4 1(V.) (VI. )Owing to their importance in connexion with the production ofvat colouring matters, many new polycyclic quinones have beenprepared m d described. They do not call for any special mentionexcepting perhaps the compound (IX), t o which the name anthan-throne has been given.61 The structure of this substance is placedbeyond dispute by the fact that it is prepared both from l:l’-di-napht’hyl-8 : 8’-dicarboxylic acid (VII) and from 1 : 1’-dinaphthyl-2 : 2’-dicarboxylic mid (VIII) by the dehydrating action of sul-phuric acid:/\/\ /\/\ /\A.\/\/ \/\/\y) \/\/ I l l I l l 1 1 ICO,HH0,C C O P --+ ():I I I.f- I/\/\ \/‘/\ HO,C, /\/\ I , I l l I l l\/\/ \/\/ \/\/(VlI.) (IX.1 (VIII.)13enshydroZ.-A rQsum6 of the chemistry of this substance isgiven by J. Sabatier and M. Murat,62 who have attempted toprepare it by the action of water on the Grignard compound,CHPh,-OMgBr. They obtained, however, a yield of only 3 percent., the main products being diphenylmethane and s-tetraphenyl-ethane; since benzophenone is also formed, i t is assumed that thereaction is represented by the equations :3CHPh2*O€I = 2H20 + COPh, + CHPh2*CHPh,.2CHPh2*OH = R20 + COPh, + CH2Ph2.I n a later paper the same authors63 show that a nearly theoreti-cal yield of benzhydrol can be obtaine’d by the action of benz-aldehyde on magnesium phenyl bromide.It is of interest to note that benzhydrol derivatives containing asubstituted amino-group in the para-position are split by bromine,6o A.Kotz and E. Schaeffer, J. pr. Chent , 1913, [ii], 88, 604 ; A., i, 186 ;J. D. Riedel, D.R.-P. 266405 ; A., i, 186.L. Kalb, Ber., 1914, 47, 1724; A , i, 849.62 Compt. rend., 1913, 157, 1496 ; A., i , 168.m Ibid., 1914, 158, 534 ; A ., i, 404.1 116 ANNUAL REPORTS ON THE PROGRESS OF CHEMISTRY.forming aldehydes and bromo-substitut,ed anilines.64 The reaction isprobably represented in the follo,wing manner :OH*CHPh*C,H,*NR, -+ O€€*CHPh*C6H4*NR2Br, -+OH*CHPh=NR,Br-C,H,Br -+ CHPhO + C,H,Br*NR2,HBr.Rupture of the Benzene Ring without Begmdcction.-The ruptureof the benzene nucleus which is illustrated by the formation ofacetylacrylic acid by the action of fuming sulphuric acid on 3-nitro-pcreso165 has been re-investigated by H. Pauly, R. Gilmour, andG. Will.GG They find that the acid C7H804, which is produced inthis reaction. is not acetylacrylic acid, but P-methyl-y-crotonolac-tone-y-acetic acid :QH: C hfe co---0 >CI€*CH,*CO,H.The course of the reaction is merely hydrolytic ; 3-nitro-pcresolreacts, in the first.instance, in its isomeric form, that is, as thenitrolic acid, and is hydrolysed t o the hydroxamic acid,CO,H*CH:CH*C?M&CH*C(OH):NOH.This is then further hydrolysed to hydroxylamine and P-methyl-inuconic acid, which subsequently passes into the lactone and water.The nitrogenous by-product, which is also formed,67 and has thestructure C7H,0,N, is regarded as 2-hydroxy-4-methylpyridine-6-carboxylic acid, and is formed from the hypothetical hydroxamicacid mentioned above by loss of water and ring closure.The Fission of Tertiary Bases.-The fission of tertiary bases a tthe nitrogen atom by cyanogen bromide which was studied by J. v.Braun,68 showed that the reaction was most marked when theamine contained a benzyl group.M. Tiffeneau and K. Fuhrer c9have now investigated the fission of bases by means of acidanhydrides and acid chlorides, and find that the fission can only beeffected in those cases in which the base has the structureAr*CH,-NRR’. The action is attributed to the intermediateformation of compounds of the type:A r- CH,,:Me CO -0':i)NAcRRlwhich undergo fission on heating in the manner shown by thedotted line.64 G. J. Essclen, jun. and L. Clarke, J. Amr. CJ~enz. Soc., 1914, 36, 308 ;65 G . Schultz and 0. Low, Ber., 1909, 42, 5T7 j A . , 1909, i, 222.66 AnnnZerL, 1914, 4403, 119 ; A., i, 485.67 G. Schultz arid 0. Low, Ber., 1910, 43, 1899 ; A., 1910, i, 552.68 B e y . , 1900, 33, 1438, 2728, 2734, 2965 ; A., 1900, i, 430, 641, 687.A., i, 278.Bull.Xoc. chim., 1914, [iv], 15, 162 ; A., i, 517ORGANlC CHEMISTRY. 117Carbon-ring Formation,A systematic study of this important question has been started,;Oand it is evident that we have still much to learn respecting theconditions which control t h e formation of cyclic compounds fromopen carbon chains. It is certain that whereas the Baeyer" Spannungstheorie " serves, and has served, as a good workinghypothesis, it is merely that and nothing more, since the tendencyto ring-formation exhibited by carbon chains depends not only onthe angle of the particuIar ring, but also to a very considerablee x t a t on the condition 'of the carbon atoms forming the chain.An admirable summary of some of the more salient facts, and alikely explanation is given by J.Kenner,71 but it is evident thatmuch work remains to be done before this question can be com-pletely answered. There can be no doubt that the presence in theopen-chain compound of a carbon atom which, in the resulting ringcompound, will be quaternary, has a marked influence against thetendency to ring-formation. There are numerous examples illus-trating the truth of this statement, and perhaps the simplest andbest is that given by L. J. Goldsworthy and W. H. Perkin,72 whopoint out that the yield of ethyl cyclopropane 1 : l-dicarboxylate (I)from ethylene dibromide and the sodium derivative of ethylmalonat'e is very small, whereas ethyl cyclopropane-1 : 2-dicarboxy-late (11) is readily prepared in good yield by a similar method:It follows, therefore, that rings containing a quaternary carbonatom ar0 relatively less stable than those in which a carbon atomof this kind is absent.This point is well brought out by Kenner,73and is also illustrated by the two compounds mentioned above,because, whereas ethyl cyclopropanel : l-dicarboxylate (I) is readilysplit by hydrogen bromide, the ester (11) is very stable towards thisreagent. It is interesting t o note that still another example of thiskind has been recorded during the year, namely, that given byA. Haller and R. Cornubert74 in the production of aa6-trimethyl-hexamide (IV) from 1 : 1 : 3 : 3-tetramethylcyclopentan-%one (III)by the action of sodamide:?H2*c'1e2>C0 + CHMe2*CH,*CH,*CMe,*CO*NH2.c H 2- c l\il e270I174(111.) (1 V.)L. J. Goldswortliy and W. H. Perkin, jun., T., 1914, 105, 2665.T., 1914, 105, 2685.Compt. rend., 1914, 158, 298 ; A . , i , 291.V2 LOC. cit. 73 LOG. cit., p. 2689118 ANNUAL REPORTS ON THE PROGRESS OF CHEMISTRY.Cyclic Compii~nds from Ethyl GZutaconate.-The condensationc.f t'wo molecules of ethyl glutaconate under the influence of con-densing agents has been studied by E. E. Blaise75 and by H. v.Pechmann.76 The first-named, using sodium ethoxide a t 100°,obtained a compound which he considered t o have formula (I),whereas v. Pechmann and his collaborators, working in etherealsolution, obtained a compound to which either formula (11) or(111) was assigned :FH,*CO* CH (CO,Et)*CH : CH*CO,EtCH :CH*CO,E t(1.)CH(CH2*C0,Et)*C(C0,Et)>cHCH,<(-*--- CH( CO, E t)(11.)C B (C H, * COPE t) C (C02E t)>CH C02Et*CH<C3 ---CH,(111.)The two compounds melted a t much the same temperature, butapparently differed in some of their properties.It is now found77that the compounds prepared by the two methods are identical, andthat, of formulze I1 and 111, the latter is the more probable.The format'ion of ring structures by the spontaneous decomposi-tion of acid chlorides, which was originally investigated by W.Borsche and his co-workers,7~ has been extended79 to the prepara-tion of hydrindone derivatives. An example of this method, whichis of importance since it renders the use of aluminium chlorideunnecessary, may <be given in the formation of methylenedioxy-a-hydrindone (V) by the distillation of piperonylpropionyl(IV) :\ / \ A H 2 ++c*2(V.1CH2(IV.1chlorideHCI.It is sbated that, contrary t o the views expressed by others,*(' theelimination of hydrogen chloride in this manner is not greatlyinfluenced by the presence of phosphorus compounds.It is well known that the five-carbon ring is usually formed with75 E. E. Blaise, Compt. rcnd., 1903, 136, 639 ; A . , 1903, i, 400, 548.7~3 H. v. Pechmann, W. Bauer, and J. Obermiller, Ber., 1904, 37, 2113; A . , 1904,77 R. Curtis and J. Kenner, Y., 1914, 105, 282.78 Ber., 1911, 44, 2942 ; A . , 1911, i, 1018.79 W. Borsche arid W. Eberlein, zbid., 1914, 47, 1460 ; A., i, 699.i, 592.H. Leuchs, J. Wutke, and E.Gieseler, ibid., 1913, 46, 2200 ; A., 1913, i,855 ; H. Lecher, ibid., 2664 ; A . , 1913, i, 1166ORGANIC CHEMISTRY. 119the greatest ease and in the presence of the mildest reagents, butthat slight changes in the structure' of the1 open-chain compoundsometimes either hinder or entirely prevent the closing of thering. An instance of this kind is recorded by E. E. Blaise,81 whofinds that s-dipropionylethane (VI) is readily converted intol-met~hyl-2-ethyl-A~-cyclopenten-6-one (VII) by 10 per cent. aqueouspotassium hydroxide :CH,* CH, CH,*CH2I >co + 1 >coCH,*CH,-CO CH2*CH, C'H,*CH,*C==C*CH,(VI.) (VII.)whereas acetonylacetonel and acetoiiylacetophenone are quite un-changed under these conditions.Terpenes and Allied Compounds.Campheiiic A cid.-The structure assigned to this acid byAschan a2 has been verified by its synthesis.83 Ethyl cyclopentan-l-one-3-carboxylate (I) is condensed with ethyl a-bromoisobutyrate inthe presence of zinc, when ethyl 3-carbethoxycyclopentan-1-01-1-iso-butyrate (11) is produced.ThO elimination of water from thissubst,ance by the aid of potassium hydrogen sulphatel gives theethyl ester (111) of either 3-carboxy-A]- o r Akyclopentene-1-iso-butyric acid, and when the free acid from this is reduced byhydrogen and platinum black, dl-cis-camphenic acid, that is, 3-carb-oxycyclopentane-1-isobutyric acid (IV), is produced :7H2-- CH2>C0 + 7H2-- H 2 > ~ (0 H) c ~ e , . CO,ECH(C0, E t) C H, CH(CO,Et)*CH,(1.1 (11.)y 3 2 - Crf2>C*CMe2* C0,EtCH(CO,Et)*CH+ I CH,--- CH2>C€€*ClA!Ie2*C0,HCH(C0,H) CH, or7H2--- CH>C*C31e,* CO, E t C H( C0,E t) CH,( I JI.) (IT.1The synthesis of camphenic acid in this manner removes the lastdoubt as to the correctness of the Wagner formula for camphene.Wagner himself 84 suggested that the remarkable change from theborceol t o the camphene structure could be regarded as similar tothat which occurs in the pinacolin rearrangement. Meerwein g581 Compt. Tend., 1914, 158, 708 ; A., i, 546.s2 B?maZen, 1910, 375, 336 ; 1911, 383, 52 ; A., 1910, i, 709 ; 1911, i, 797.P. Lip", Ber., 1914, 47, 871 ; A . , i, 542.84 J. h!uss3. Plzy.?. Chcna. Xoc., 1899, 31, 680.85 Aitnnht, 1914, 405, 129 ; A., i, 850120 ANNUAL REPORTS ON THE PROGRESS OF CHEMISTRY.CH-has now brought forward evidence which shows that this explana-tion is correct.He points out that the transformation of borneolinto camphene can be represented thus:(JJ CH (4) CH-CH, CH,-C-UH*O*SO, CH,*C-CH-OORGANIC CHEMISTRY. 121salt of this acid is distilled in a current of carbon dioxide r-fencho-camphorone is produced :YH,*YH*CH2*C0,H 7 H2* 7 K--yH2C'H2*CHoC02H UH,*CH--CO1 p e 2 --f I t)Me2IHydrocarbons Allied to the Terpenes.-A method has beendescribed 89 by which the action of the Grignard reagent on iiitrilesmay be utilised for the production of certain hydrocarbons alliedto the terpenes. Thus the condensation of 1-methylcyclohexan-3-one(VIII) with the sodium compound of ethyl cyanoacetate leads, inpart, to the formation of the cyano-acid (IX).Elimination ofcarbon dioxide yields the nitrile (X), from which by the action ofmagnesium methyl iodide the ketone (XI) is produced. Furthertreatment with the Grignard reagent then yields the terpineol(XII), which gives the hydrocarbon (XIII) on dehydration :CHMe*CH>C*CH,*CMe2*OH CH,<CH2-CH, CHRle*CH>c. CH: cJ$e2CH2<CH,-CE€2(XII.) (XIII.)Benaoterpenes.-A series of experiments has been started 90 withthe object of synthesising substances having a similar structure tothe terpenes, b u t containing an aromatic residue (XIV). It isthought that a study of these substances will help to eIucidate theproblem of the structure of the sesquiterpenes, since the relation-ship between naphthalene and the hydroaromatic sesquiterpenes isshown by the following formulze:5=3 7H3CH CH QH CH CH2 CCH/\C/\CH CH/\C/\FH, C H , * C H / \ ~ / \ C EI ICH,CH\j/h\#X€ CH CH CH(\,k!\,CH2 CH $!H CH2\/CH\/ CH,CH(CH,), CH,*C:CH,Naphthalene. Ben zo-p-men thane.a-Selinene.(XIV. )39 W. N. Haworth and A. W. Fyfe, T., 1914, 105, 1659.9O F. W. Kay a i d A. Morton, ibid., 1565122 ANNUAL REPORTS ON THE PROGRESS OF CHEMISTRY.It is open t o question whether any useful purpose is served bygiving the name henzo-p-menthane to a substance which is obviouslya h u e derivative of tetrahydronaphthalene.Various Reactions.The tendency to the formation of hydroxy-anhydrides (6-hydroxy-a-pyrones) from glutaconic acid and some of its derivatives, whichwas described by N.Bland and J. F. Thorpe,gl has been shown byW. DieckmannQ2 to be exhibited also by homophthalic acid (11).A comparison of the hydroxy-anhydrides of homophthalic acid andglutaconic acid (I) shows that both yield yellow neutral salts withalkali, but that the anhydride from homophthalic acid gives nocolour with ferric chloride and does not absorb bromine. I n thefree state it has, therefore, the normal structure (111):CH CH(1.1 (11.) (111.)Homophthalic acid is therefore comparable with, for example,aP-dimethylglutaconic acid 93 (IV), which, although it yields ananhydride that dissolves in alkali, does not give a coloration withferric chloride, and thus according t o Dieckmann should, in thefree state, have the structure V I :,CH,*CO,HCH,$CH,*C\CO,HCHCOCH,*C 0co \/(IV.) (V.) (VI.1There is, of course, a fallacy in Dieckmann’s argument, for if theanhydride of glutaconic acid has the enol structure because i t givesa yellow sodium salt and a coloration with ferric chloride, why isthe free anhydride colourless? This error is apparent in a greatdeal of the work that has been done recently on keto-enol tauto-merism. I t has been too hastily assumed that, because the ferricchloride and bromine-absorption reactions serve as excellent meansfor distinguishing between the keto and enol forms of certain types,therefore, when compounds fail to respond t o these tests, theabsence of the enol modification is proved. It must be remembered91 T., 1912, 101, 863.93 F.B. Thole and J. F. Thorp, T., 1911, 99, 2216.92 Ber., 1C14, 47, 1428 ; A . , i, 690ORGANIC CHEMISTRY. 123that, of the two isomeric phenols thymol and carvacrol, onlycarvacrol gives a coloration with ferric chloride, and there are,O HCarvacrol.OHThymol.moreover, numerous other examples which show that whereas apositive ferric chloride reaction is a sure indication of the presenceof the enol variety, a negative result does not necessarily showits absence.Reduction of the Carboizyl Group.-The valuable method for thereduction of aldehydes and ketones which was introduced by E.Clemmensen94 has now been improved and extended.95 The methodpromises to be of special service in the preparation of homologuesof the various phenols from hydroxy-ketones, and although its valuefor the reduction of hydroxy-aldehydes is diminished by the sensi-tiveness of many of these compounds towards hydrochloric acid, themethyl-substituted phenols obtained are of a high degree of purity.The ketones are reduced by amalgamated zinc in the presence ofhydrochloric acid, and the applicability of the method is illustratedby the production of numerous substituted phenols from the corre-spofiding ketones ; thus p-hydroxyacetophenone yields p-ethyl-phenol, m-acetyl-o-cresol gives 2-methyl-4-ethylpheno1, acetylquinolis reduced to ethylquinol, and so forth.Bsterification b y Ultm-violet Light.-It has been noticed duringthe course of certain experiments having f o r their object the trans-formation of stereoisomerides by ultra-violet light,g6 that alcoholicsolutions of certain acids undergo esterification when exposed tothese rays.Thus benzoic acid is esterified to the extent of 30 percent. in eight days, and t o the extent of 56 per cent. in the sametime when a trace of hydrochloric acid is present.Phototropy and Thermotropy.-That trituration or rubbing iscapable of producing polymorphic change and is an importantmeans of bringing about chemical reactions is well known, and thepaper recently communicated by L. H. Parker97 not only adds toour knowledge of this subject, but also contains a useful list ofreferences to previous work. The principle has now been applied98to a number of 4-hydro~ybenzylidenearylamines~ the general94 Ber., 1913, 46, 1837 ; A ., 1913, i, 733.95 E. Clemmensen, ibid., 1914, 47, 51 ; A., i, 271.96 R. Stoermer and H. Ladewig, ibid., 1803 ; A . , i, 966.97 T., 1914, 105, 1504.98 A. Senicr and R. B. Forster, ibid., 2462124 ANNUAL REPORTS ON THE PROGRESS OF CHEMISTRY.results of which may be described by referring to those obtainedwith 4-hydroxybenzylideneaniline, OH*C,H,*CH:N*C,I3,. Thissubstance is dimorphous ; trituration changes it into a deeper-coloured dimorphic variety, which? however, on long keepingreverts t,o the original compound. The paler-coloured variety isthermotropic, becoming paler a t ‘‘ the lower temperature ” (that is,of solid carbon dioxide9, and returning to the original colour a t theordinary tlemperature. The deeper-coloured variety is not photo-tropic, but on prolonged exposure to sunlight changes to a brown,polymorphic form.J.F. THORPE.PART III.-HETEROCYCLIC DIVISION.IN former years, the task of the reporter was rendered difficult bythe great mass of material from which it was necessary to selectthose subjects which lent themselves best to summarisation inbrief; but in the current Report this embarrassment has not beenso great as usual, f o r only the first seven months of the year’s workare covered by the Austrian and German journals in our possession,so that the amount of published matter to be dealt with isless than in any previous year. This has necessitated a choicebetween two policies, for it was possible either to make the Reportof the customary length by including various pieces of work ofless importance, or t o reduce thO length of the Report to someextent and keep its standard approximately the same as in othervolumes.The latter objective has been chosen, as it seems themore desirable of the two, so that the present Report remainscomparable with its forerunners.Last year saw the close of the main investigations on the con-stitution of chlorophyll, and in the period covered by this Reportit cannot be said that much more light has been thrown on thebroad outlines of the problem. The work of the year in this par-ticular field has taken a fresh turn, and t.he researches of variousinvestigators have been directed towards the occurrence of chloro-phyll in various plants, and the reactions of chlorophyll quachlorophyll, rather than those of its decomposition products.I nfact, it may be said that chlorophyll a t present is falling backfrom the centre of the chemical stage and giving place to othersubjects of interest.The same holds good with regard t o the chemistry of the bloodpigment. Great advances in our knowledge of this substance weremade in 1913, but the last twelve months have left the main quesORGANIC CHEMISTRY. 125tion very much where i t was, and have been spent in that slowand laborious work which marks the accumulation of data thatare necessary before a further advance is possible. This amassingof isolated facts does not lend itself to reproduction in connectedform, so that in the present Report only a brief mention is madeof some points; but it must be borne in mind that; those fieldswhich have been selected for treatment represent only a smallfraction of the work accomplished in the twelvemonth, and it mustnot be assumed that the investigation of haemin is approaching astandstill.Haemin and chlorophyll having thus passed into a new stage oftheir chemical career, we turn to, another field to find the subjectwhich promises to take their place.From the developments whichhave recently occurred, it appears likely that this will be foundin the study of the pigments of plants and flowers. I n fact, as faras the heterocyclic series is concerned, this year has been markedby a greater advance in this direction than any which have takenplace in other branches of the subject.Willstatter and hiscolleagues have not only broken new ground, but have alreadyestablished the main lines along which further advances can bemade, and, what is even more encouraging than the success of asingle school, many other workers have been drawn into the fieldin questmion, so that i t appears certain that our knowledge of itwill soon progress both extensively and intensively.These developments in the study of chlorophyll, hzemin, and theplant pigments appear to mark the recrudescence of organicchemistry proper, in contradistinction from what may be termedthe chemistry of the carbon compounds. Thus pure chemistry andphysiological chemistry are again tending t o become more closelyallied than they have been since the earliest days of modernchemical history, and this approximation appears one of the mosthopeful signs in present-day science.Under the conditions exist-ing in living organisms, many reactions take place a t the ordinarytemperature which in the laboratory can only be brought aboutby the use of active reagents and the application of heat, and itseems possible that the coming years may see determined attemptsto reproduce these reactions under conditions more similar to thosewhich suffice for the syntheses of compounds within vital tissues.Turning now to fields of less general interest, the investigationsof haemin and chlorophyll have necessitated a closer investigationof the pyrrole group, and a considerable amount of work has beencarried out in this branch of the subject during the year, althoughmuch of it is too det'ailed to summarise in these pages.The chemistry of the alkaloids has not been marked by an126 ANNUAL REPORTS ON THE PROGRESS OF CHEMISTRY.great advances, but rather by a steady accumulation of know-ledge upon minor points, which will serve its purpose later on.New syntheses of coumarin and thiophen have been devised, ofwhich the latter is important, as it permits of the substance beingobtained in quantity with ease, and it seems likely our knowledgeof the thiophen group may soon increase considerably.Anotherpoint t o which attentmion might be drawn is the discovery thatpyridine acts on many sulphur compounds catalytically, so that itis a solvent that must be utilised with caution for substances ofthis type.Surveying the heterocyclic group as a whole, it cannot be saidthat the year has been by any means an unsatisfactory one asregards research in this division of the subject.Cyclic Compounds the Rings of which coiztain Mercury Atoms.The number of elements which have hitherto shown a tendencyto act as members of heterocyclic rings is comparatively limited, andi t is therefore of some interest to find that this year has brought anincrease in this class.Not only do we find mercury atoms play-ing the part of ring-members in the case of substances such asN H,OAcso, -in which the heterocycIic ring is produced by a kind of internalsalt-formation, akin to the betaine or lactone structure, but com-pounds have been synthesised which are mercury analogues ofpiperidine or penthiophen, and in some cases more than onemercury atom occupies a place in the ring.1The preparation of the dimercuri-compounds begins with the pro-duction of a Grignard reagent from m-dibromopentane.Themagnesium compound is then treated in ethereal solution with dry,powdered mercuric bromide, and by this means penbamethylene-a€-dimercuridibromide is quantitatively formed. I n this coni-pound, the mercury atoms are very firmly attached to the rest ofthe molecule, whilst the halogen atoms are extremely reactive, sothat the corresponding iodide, nitrate, or hydroxide can be obtainedfrom the dibromide.The dibromide is also attacked by either hydrogen sulphide o racetylene, yielding compounds the simplest formulz of whichwould beR.Hrieger and W. Schulem~iin, J. pr. Cheqn., 1914, [ii], 89, 97 ; A . , i, 611ORGANIC CHEMISTRY. 127but as the molecular weight of these has not been ascertained, itis possible that they are polymerides.2The bromine atoms are also replaceable by dibasic acid radicles,and in this wag rings containing a very large number of atomscan be produced; for example, azelaic acid yields an eighteen-membered ring.The action of magnesium phenyl bromide on the mercuribromideresults in the production of pentamethylene-aedimercuridiphenyl,Ph*Hg* [CH,],*HgPh, a representative of a mixed mercury dialkyl.Turning now to the problem of preparing cyclic compounds con-taining only a single mercury atom in the ring, it cannot yet besaid that this problem has been placed beyond doubt, for themolecular-weight determinations seem t o have been unsatisfactory,so that we cannot be certain that the substance produced has theformula I and not 11:The compound was prepared by the action of sodium amalgam ona€-dibromopentane. It appears t o polymerise readily, thus show-ing an analogy to some cyclic compounds containing oxygenand sulphur atoms in the ring.Its most characteristic reaction isshown when it is treated with iodine, dihalogenopentanes andpentamethyleneaedimercury haloids being formed.Thiopken and some of its Derivatives.Hitherto the synthesis of thiophen has been a somewhatlaborious one, but a new process has recently been discovered3whereby it, may be produced in large quantities with the minimumof trouble.Pyrites is heated at about 300° in a tube throughwhich acetylene is passed, and i t is found that about half theproduct is thiophen, which can easily be purified from theby-products of the reaction. The essential condition for success isthe employment of the pyrites in a very finely powdered state.When the apparatus is working properly, it is possible to preparemore than 300 grams of thiophen per diem.The substitution of isoprene or Py-dimethyl-il"y -butadiene foracetylene gives rise to 3-methylthiophen and 3 : 4-dimethylthiophenrespectively, but the yields are very poor.4The nitration of thiophen has been studied, and it is found thatwhen t,hiophen dissolved in acetic anhydride is nitrated at 0-5Oby means of nitric acid (D 1-52) dissolved in acetic a.nhydride (airS. Hilpert and G.Griittner, Ber., 7914, 49, 177, 186 ; A., i, 261, 262.W. Steinkopf and G . Kirchhoff, An?zaZcn, 1914, 4103, 1 ; A . , i, 425.TIr. Steinkopf, ibid., 17 ; A . , i, 426128 ANNUAL REPORTS ON THE PROGRESS OF CHEMISTRY.being drawn through the mixture during the operation), the pro-duct is 2-nitrothiophen in a 70 per cent. yield.From the nitrothiophen thus formed, the corresponding amino-thioqhen can be obtained by reduction with tin and hydrochloricacid. It is not very stable, for in the presence of a trace ofoxygen it polymerises, and subsequently undergoes oxidation.The action of mercuric chloride on thiophen gives rise t o bothmono- and di-mercurichlorides. Investigations have now been seton foot to discover the influence on the reaction which is exertedby substlituents in the thiophen nucleus.5 When both a-positionsof the thiophen ring are free, it is possible to obtain a dimercuri-derivative as well as the mono-compound, but if one a-position beoccupied by a substituent, then, as a general rule, only the mono-mercurichloride is produced.When both the a-positions in thethiophen ring are substituted, the results of the reaction aredifferent, for in some cases the end-product is an additive compoundcontaining a mercurichloride group in position 3, whilst in otherinstances no interaction a t all takes place between the substitutedthiophen and the mercuric chloride.These rules are evidentlysubject to exceptions, and do not furnish us with a trustworthymethod of differentiating between isomeric thiophen derivatives.Thiophen-2-mercurichloride reacts with soldium in boiling xyleneto form mercury 2 : 2/-dithienyl, Hg(C,H,S),.This work has led to an interesting discovery of a method bymeans of which thiophen ketones can be prepared. Thiophenis allowed to react with mercuric chloride, producing thiophen-2-mercurichloride. By the action of acetyl chloride on the lattersubstance, 2-acetylthiophen is formed and mercuric chloride iseliminated. I n this way, a small quantity of mercuric chlorideshould suffice t’o convert a large quantity of thiophen into theketone, and a reaction somewhat akin to the Friedel-Crafts’synthesis has been discovered.The yields obtained are good, andthe reaction appears t o be generally applicable.A compound of some interest has been synthesised which con-tains a thiophen nucleus condensed with a pyridine ring.c Themethod of preparation is a modification of Skraup’s quinolinesynthesis, 2-aminothiophen being substituted for aniline and2-nitrothiophen being used instead of nitrobenzene. The yield isa poor one, 5-7 per cent. of pyridino-2: 3-thiophen (I) beingDroduced :(1.15 W. Steinkopf and M. Bauermeister, Annalen, 1914, 403, 50 ; A . , i, 427.W. Steiiikopf and G. Liitzendorf, ibid., 45 ; A., i, 432ORGANIC CHEMISTRY. 129The substance is a yellow liquid resembling quinoline in its odourand in some of its reactions.A Sywthesis of Coumarin.When a mixture of o-chlorobenzylidene chloride, acetic acid, andpotassium acetate is heated a t 210-220°, the product of the re-action is o-chlorocinnamic acid.This compound, by electrolyticreduction, can be converted into o-chloro-P-phenylpropionic acid,which, in turn, by heating with sodium hydroxide and water a t250°, is changed into melilotic acid (I) :0(1.1 (11.)Distillation of this substance produces the corresponding internalanhydride (11), and by passing bromine vapour over the last-named substance a t a temperature of 270-300°, two atoms ofhydrogen are removed in the form of hydrogen bro,mide, andcoumarin is produced.'The Condensation Products of PhloroglucinoE and Aldehydes.When three molecular proportions of phloroglucinol are con-densed with two of an aldehyde, two molecular proportions of waterare eliminated, but a careful examination of the products showsthat in actual practice two reactions are taking place side by side,with the production of two different end-products.* The state ofaffairs may be expreesed in the following manner :Reaction I.1 mol. Phloroglucinol + 1 mol. Aldehyde - 1 mol. Water.,, 11. 2 mols. ,, -1-1 mol. ,, -1 mol. ,,Total effect. 3 mols. ,, +2 mols. ,, - 2 mols. ,?The course of the condensation is much affect'ed by the conditionsemployed and by the nature of the condensing agent used. Theinitial condensation products are colourless and amorphous, but ifthey are kept in contact with the mother liquor a further elimina-tion of water takes place, and red compounds are obtained.Thuswhen /3-hydroxybutyraldehyde is condensed with phloroglucinol bymeans of dilute sulphuric acid a t about 6O, the product has thecomposition C,,H1204, and appears to have the structureH. Meyer, R. Beer, and G. Trasch, Monatsh., 1913, 34, 1665 ; A . , i, 44.8 I?. Wenzel, ibid., 1915 ; A , , i, 75.REP.-VOL. XI. 130 ANNUAL REPORTS ON THE PROGRESS OF CHEMISTRY.This substance is deposited from the solution in snow-white flocks,which turn to a pale yellow powder on drying; but, if i t is allowedto remain in contact with sulphuric acid for a t h e , it loses asecond molecule of water and yields a red substance, which appar-ently has the structure represented byThe reaction appears to be a fairly general one, being applic-able even in the case of ths monosaccharoses and phloroglucinol.Thus I-xylose can be condensed with dimethylphloroglucinol inpresence of hydrochloric acid t o form a deep brownish-red sub-stance, which appears to have the structure represented bySome of thesesalts with acids.compounds appear to yield well-defined oxoniumDerivatives of Dime t hylpyrone.Owing to the ease with which dimethylpyrone hydrochlorideappears to become hydrolysed when i t is dissolved in water, thereseems to be some doubt as to whether it8 aqueous-solution actuallycontains the true salt of the base or merely a mixture of the baseand acid.To throw light on this question, the concentration ofthe chloride ion has been determined in solutions of differentstrengths by measuring the potential of the electrode Hg I solidHg,Cl,, solution of dimethylpyrone hydrochloride, against aN / 10-calomel electrode.Similar measurements were made withvarious solutions of hydrochloric acid. The equivalent conductivi-ties of all the solutions were also measured, and on plotting thevalues of the chloridion concentrations as abscissa: against the equi-valent conductivities as ordinates, it was found that the curve fordimethylpyrone hydrochloride lies below that f o r hydrochloric acid,both curves approaching the same value f o r infinite dilution. FroORGANIC CHEMISTRY. 131this it is deduced t'hat dimethylpyrone hydrochloride exists insolution as a true salt, although it is hydrolysed to sonie extent.$1)iniethylpyrone forms a large number of additive products withacids, phenols, and nitrophenols.10 By means of the cryoscopicmethod, the existence and nature of these substances has beenascertained, and it is found that they may be divided into threeclasses, having respectively the compositions :C7H,02,HX 2C7H,0,,3HX C71'I,0,,2HXThe additive compounds are well crystallised, and as a general ruletheir melting points are lower than those of their components.Thereaction between the dimethylpyrone and the &her reagent appearst o be ionic, and the additive compounds seem to be true oxoniumsalts.Some arylidene dehatives of dimet(hy1pysone 11 have been pre-pared which form intenseIy coloured salts with acids.It is con-cluded that acidic or weakly basic pyrones possess structurescontaining a bridge formed by the oxygen of the carbonyl group, assuggested by Collie, whilst the arylidenepyrone derivatives areassumed to possess the ordinary unbridged pyrone structure.Iitdiyo and its Allies.A considerable amount of work has been carried out on the indigoderivatives, and some reference must be made to the more importantbranches of this subject.New syntheses of indigotin12 have been devised, and methods ofpreparing substituted indigotins have been studied.13 It has beenshown that experimental conditions have a marked effect on theaccuracy of titration of indigotindisulphonic acid with potassiumpermanganate.14 The amount of the permanganate requiredappears to be about 12.6 per cent.less than the theoretical quan-tity. Efforts have been made t o track the source of this error,and i t has been found that an addition of manganese sulphatediminishes the difference between theory and practice, whilst thepresence of a large excess of sulphuric acid also helps to eliminatethe error. Almoet theoretical results are obtained when the0 13. N. K. R$rdarn, Oversigt. K. Dnnske. Yidensk. Sslskabs. Forhandl., 1914,lo J. Kendall, J. Amer. Che?n. Xoc., 1914, 36, 1222 ; A., i, 858.l1 A. A. Boon, K. J. McKenzie, aud J. Trotter, P., 1914, 30, 205 ; A. A. Boon,la W. Madelung, Annalen, 1914, 405, 58 j A , , i, 737.1Y A. Reiesert, Ber., 1914, 47, 672 ; A., i, 432 ; E. Grandmougin and P.Seyder,ibid., 2365 ; A., i, 1142 ; H. Levinstein, J. SOC. Clmn. Ind., 1914, 33, 574 ; A , , i,876 ; M. Tschilikin, J. Russ. Phys. C'hmt. Soe., 1913, 45, 1834 ; A , , i, 191.K 2243 ; A., i, 1173.F. J. Wilson, and I. M. Heilbron, T., 1914, 105, 2176.l4 H. E. Wagner, J. pr. Chem,, 1914, [ii], 89, 377 ; A. i, 875132 ANNUAL REPORTS ON THE PROGRESS OF CHEMISTRY.indigotin solution is added drop by drop to a feebly acid perman-ganate solution, whilst rapid addition results in too great aquantity of permanganate being used.It has been found that l-substituted isatins can be obtained bythe intramolecular condensation of oxamyl chlorides of the typeCl*CO*CO*NRPh; in some cases the reaction can be broughtabout by the direct interaction of oxalyl chloride and an amine inthe presence of aluminium chloride.Thus oxalyl chloride andethylaniline produce l-ethylisatin.16Flavones and Flavonols.The interaction between phenols and ethyl acetoacetate canfollow two very different linee according t o the conditions underwhich i t takes place. On the one hand, coumarin derivatives areobtained, whereas if phosphoric oxide is used instead of sulphuricacid the main products are chromones.16 The latter reaction hasnow been applied to the preparation of several substances whichhad previously been synthesised by other methods. I n the caseof the condensation of resorcinol with methyl methylacetoacetate,no difficulty was experienced in preparing 7-hydroxy-2 : 3-dimethyl-chromone. When the reaction was carried out with ethyl aceto-acetate and phenol, however, no corresponding end-product wasobtained.To account for this, it seemed possible that the presenceof the phenol had converted the ethyl acetoacetate entirely intothe ketonic form, and in order to avoid this, the sodium derivativeof ethyl acetoacetate was employed in the reaction instead of theester itself. The results in this case were normal, a condensationproduct being readily obtained.The way was now clear to utilise ethyl benzoylacetate in placeof ethyl acetoacetate, and the product of the reaction was flavone:0/\/OH /\/\g* C,H, I I H*-g*C6H, --+ I I + R*OH+H,O. \/\P" coRO-COThus a new and simple synthesis of this compound has been found,which may in time prove to be a general method for the prepara-tion of various flavone derivatives, whilst alkylated chromones canalso be obtained in the same manner from suitable reagents.17On the other hand, the method of synthmis discovered byl5 R.StollB, Ber., 1913, 46, 3915 ; A., i, 198.l6 H. Siinonis and P. Remmert, ibid.) 1914, 47, 2229 ; A., i, 980.l7 H. Siinoiiis and C. B. A. Lehmann, ibid., 692 ; A , , i, 424ORGANIC CHEMISTRY. 133Auwers and Muller,l* by which, for instance, the dibromide ofbenzylidene-4-methylcoumaran-3-one is converted into 6-methyl-flavonol by treatment with hot alcoholic potassium hydroxide, hasbeen proved19 to be incapable of general application. It is especi-ally unfortunate that it appears to break down in those cases wherethe production of naturally occurring flavonols might bO expected.Some work of considerable interest has been carried out in thequercetin group of dyes.20 These substances in general yield colourswhich range from yellow to brown, but i t seemed possible that, byintroducing suitable auxochrome groups into the molecules, adeepening of the tints might be effected which would provide red,violet, or blue colours having the fastness of the original dye.The choice of the auxochrome groups suit'able for producing therequired result is not so simple as i t appears a t first sight.I n thealizarin seriea, the hydroxyl radicle has proved a valuable auxo-chrome, but its effect in the flavone group does not appear to be sopowerful, for there is very little change in tint in passing fromluteolin through quercetin t o myricetin, although the second com-pound contains an extra hydroxyl radicle, and the last one hastwo more hydroxyl groups than are present in luteolin.Neverthe-less, the endeavour was made to introduce an additional hydroxylgroup into the quercetin molecule, for the possible results extendedinto more than one field. I n the first place, the action of an auxo-chrome is influenced by its position in the molecular structure,so that i t seemed quite likely that if the fresh hydroxyl radiclebecame attached to the ring in certain positions, it might exert amore marked effect than had been observed in other cases. Further,the synthesis of a new hydroxyquercetin might help to throw lighton the structure of the other hydroxyquercetin derivatives such asmyricetin, gossypetin, o r quercetagin.The first steps were taken by the complete methylation of thequercetin molecule with a view to protect the hydroxyl radicles init.Thereafter, the methoxylated substance was nitrated, the nitro-group reduced, and the compound finally demethylated. I n thisway 6/-aminoquercetin was obtained. Examination of its propertiesshowed that as a dye it did not differ much in tint from quercetinitself, so that the auxochromic action of the amino-group was evi-dently insufficient to produce the desired result. An attemptwas then made to diazotise the amino-compound, but it was notfound possible to displace the amino-group by hydroxyl.A curious point came to light, in the course of the investigation.lx K.v. Auwers and I<. Miiller, Ber., 1908, 41, 4233 ; A . , 1909, i, 45.19 K. v. Auwers and P. Pohl, Annalen, 1914, 405, 243 ; A., i, 981.l3. R. Watson, T., 1914, 105, 338134 ANNUAL REPORTS ON THE PROGRESS OF CHEMISTRY.Quercetin itself yields a series of bright yellow oxonium salts, butthe progressive introduction of methyl groups appears t o renderthe stability of such salts much less marked. Thus quercetin tetra-methyl ether forms only a very unstable sulphate. On the otherhand, when a fifth inethyl group enters the molecule the power ofoxonium-salt formation reappears in full strength, for the penta-methyl derivative of quercetin gives rise to several quite stable saltswith mineral acids.These substances are bright yellow, and itseems probable that they have a quinonoid structure of some typesuch as:0 OMeI 1 ‘CmOMe\,/ \c1Meo/\/\c-/=Lo( CH,\A//Me0O HWhen exposed t o air, the salts acquire a surface tint of bright red,for the occurrence of which we have a t present no explanation.The foregoing methods having failed t o produce a deeper-coloureddye, various other attempts were carried out.21 The intermediatecompound, 6’-nitroquercetin pentamethyl ether was partly demethyl-ated, yielding 6’-nitroquercetin dimethyl ether ; but this was found t obe useless for dyeing purposes. Various attempts t o introduce addi-tional auxochromes failed, and the multiplication of chromophoricgroups in the molecule also gave rise to unsatisfactory results.The objective sought wast o introduce into the molecule some substituent which wouldproduce a permanent quinonoid structure, and this was gained byreplacing the pyrone ketonic radicle by the group *CR*OH.Thenew compounds of this type might be expected to resemble inbehaviour the dyes of the triphenylcarbinol series. Compounds ofthe following structure (I and 11) were therefore prepared by theaction of Grignard’s reagent on quercetin pentaethyl ether andsubsequent de-ethylation with hydriodic acid. The base corre-sponding with I1 dyes wool violet o r crimson, according as t owhether alum or tin is used:Success was attained in another way.OH T Il0 OEt 0EtO 7 HO 7Et Et(1.) (11.)21 E. R. Watson and K.B. Sen, T., 1914, 105, 389 ; compare R. Willstatterand H. Mallison, Sitxungsber. K. Akad. Wiss. Berlin, 1914, 771 ; A., i, 1081ORGANIC CHEMISTRY. 135Plant Pigments and Allied Materials.Last year saw the beginning of a systematic investigation of thecolouring matters of various flowers, and although the resultsobtained a t first were meagre, they have now been considerablyextended. The earliest work dealt with the pigment which givesits colour to the cornflower,22 and this compound was obtained in apure, crystalline form, which permitted a chemical investigationof its nature. The results attained tend to show that the antho-cyanins, as these pigments are' termed, belong to a class of basicveget'able, compounds which owe their basicity to an oxygen atom.The new class of pigments differs from the better known flavonedyes in that the latter form only easily dissociated salts with acids,and therefore do not occur as oxo,nium salts in the plants fromwhich they are derived; the anthocyanins, on the other hand, cancombine with acids t o yield comparatively stable oxonium deriv-atives, somewhat akin to those of phenopyrylium 23 :An examination of the cornflower pigment, cyanin,24 has thrownlight on the problem of the colour variations which are found inthe flowers.Apparently there are three main forms in which thecompounds can exist. When combined with mineral acids o r theacids of plants, the anthocyanins are red in tint. Neutralisationof the acid causes a change of colour to violet, and it seemsprobable that the violet type of compound is an internal salt akinto the phenolbetaines, and contains the group:>c-o-o<The addition of alkali produces blue alkali salts, which are to beregarded as derivatives of the neutral violet forms in which nochange has been effected in the internal oxonium salt complex.Thus, what a t first sight appear to be three distinct' colouringmatters are merely modifications of one primary structure underlocal variations in the distribution of acid and alkali throughoutthe plant tissues.The earlier investigations of these pigments led to the proof that22 R.TVillstatter am1 A. E. Everest, Annalen, 1923, 401, 189 ; A . , 1913, i, 1371.H. Decker and T. von Fellcnberg, ibid., 1908, 364, 1 ; A., 1909, i, 116 ;compare also H.Decker and P. Becker, Ber., 1914, 47, 2289 ; A., i, 1082.24 R. Willstatter, Sitxzcngsber. K. Akad. Wiss. Berlin, 1914, 402 ; A., i, 564136 ANNUAL REPORTS ON THE PROGRESS OF CHEMISTRY.the colouring matters are present in the plants in the form ofglucosides, so that this chemical class has been found in a hithertounsuspected field.Hydrolysis of the pigment of the cornflower leads t o its decom-position into two molecules of dextrose and one molecule of theactual colouring matter, which has been termed cyanidin. Thelatter substance has now been shown to have the compositionC15Hl106Cl. The previous estimate of its composition has beenfound t o be erroneous on account of the specimen not having beenentirely freed from water, which apparently is very stronglyretained by the compound.The anthocyanin which is present in the rose (rosa gallicu) isa diglucoside of cyanidin, whilst the pigment of the cranberry iscomposed of one molecule of galactose combined with one moleculeof cyanidin.From this i t will be seen that cyanidin is very widelydistributed in nature.Several other anthocyanins have been investigated. Thus grapesowe their colour t o the anthocyanin oenin; bilberries are colouredby the compound myrtillin; and, in addition to these, delphininand pelargonin have been purified.The methods of extracting the colouring matter naturally varyaccording t o the plant which is used as a starting material, but theessential part of the process is the formation of a sparingly solubleoxonium salt.I n the case of grapes, the skins are extracted withglacial acetic acid a t the ordinary temperature and the dark redfiltrate is precipitated with ether. From the deposit thus obtaineda picrate is prepared, and by this means the pigment is purified andthe anthocyanin itself obtained. When bilberry skins are workedup, the process is different. The extraction is carried out with a1 per cent. alcoholic solution of hydrochloric acid; the liquor thusobtained is precipitated with ether, and the pigment is removedfrom the precipitate by means of water. By the addition of con-centrated hydrochloric acid and cooling of the solution, the antho-cyanin chloride is precipitated in an almost pure condition.The investigation of the anthocyanins is rendered difficult bythe fact that they are not very stable substances.Aqueous oralcoholic solutions of the pigments gradually lose their colour, insome cases with great rapidity. This alteration in colour does notappear t o be due t o a reduction of the pigment, but is t o be attri-buted t o an isomeric change similar t o that which takes place whena triphenylmethane dye is converted into the corresponding car-binol. It has been found that the decolorisation can be delayed bythe addition of salts such as sodium chloride, and the addition ofexcess of mineral acid stops i t completely. Further, a cyanidiORGANIC CHEMISTRY. 137solution which has become decolorised can be restored to its originalcondition by the addition of acid.The absorption spectra of the anthocyanins closely resemble oneanother; in acid solution they all show a broad absorption band,which extends over the green and blue regions of the spectrum.Hydrolysis of the anthocyanins produces the correepondinganthocyanidins, which in general resemble each other, althoughthey differ to some extent in crystalline form, solubility, andcolour-reaction with ferric chloride.They are much more stablethan the parent anthocyanins, for they are apparently not decolor-ised on keeping in solution. On heating, however, they lose theircolour, and this seems to be due to the conversion of the quinonoidpyrone structure of the anthocyanidin into an ordinary pyroneform in which t’he bridge oxygen has lost its strongly basic proper-ties to a great extent.Turning now to the question of the constitution of these sub-stances, a certain amount of information has already been gained.An =xamination of their empirical formulce suggests a kinship withthe flavone dyes, for the anthocyanidins are isomeric with severalsubstances of the flavone series.Thus cyanidin is isomeric withluteolin and fisetin, pelargonidin is an isomeride of apigenin,whilst delphinidin, quercitdn, and morin all have the same molecularcomposition.Further relations between the anthocyanidins themselves comet o light when we examine the decomposition products which arisefrom them on heating with alkali.Anthocyanidin Deconi posi tion Prodnc ts.l’elargonidiu, C,,HloO, Phlorogluciriol and p - Hydroxybenzoic acid.Cyanidin, C,,Hlo06 2 ) ,, Protocstechuic acid.Delphinidin, ClsH,oO, , I ,, Gallic acid.These reactions help t o clear up the question of the position ofthe oxygen atoms in the various anthocyanidin molecules.I n thecase of pelargonidin, for example, there are five oxygen atoms t oaccount for; three of these are in the phloroglucinol portion, oneis in the hydroxyl group of the phydroxybenzoic acid, a fourthmust go t o form the pyrone bridge, so that there is only a singleoxygen atom left unaccounted for out of the five. Similar reason-ing holds for the other substances.Now when i t is recalled that the anthocyanins are apparentlypyrylium derivatives there are two possible ways of regarding them;they may be considered as substitution products of either 2-phenyl-pyrylium or of 4-phenylpyrylium 138 ANNUAL REPORTS ON THE PROGRESS OF CHEMISTRY.c1 c:\/On further examination of cyanidin chloride, more light wasEvidently this substance, on the above thrown on the problem.assumption, might be either (I) o r (11) :c1 CI0 OH 0I ICareful treatment with alkali revealed results which decided thechoice.25 Were (11) the correct formula, i t should be possible t oisolate as an intermediate product a substituted benzophenone[maclurin, C6H3(OH),*CO*C6H,(OH)3], but no such substancecould be found.Further, formula (I) suggests the possibility thatby the oxidat’ion of the anthocyanin a dyestuff of the flavone seriesmight be produced or, conversely, the reduction of a flavone deriv-ative might yield an anthocyanin.The latter process seemed morepromising, and quercetin was chosen as a test substance. Whenreduced in alcoholic solution by means of magnesium and hydro-chloric acid in the presence of mercury a t Oo, quercetin was foundtQ yield a substance, allocyanidin chloride, which readily loseshydrogen chloride, and is converted into allocyanidin. Thisresembles closely, but is not identical with, cyanidin. When thereaction is carried out a t 3 5 O , however, the end-product is foundto be a mixture of allocyanidin and cyanidin itself, the lattersynthetic substance being identical with the natural productderived from roses or cornflowers.2b R. WillstPtter aud H. Mallison, Sitzusbgsber.K. Aknd. Wiss. Eerlin, 1914, 769 ;A., i, 1081ORGANIC CHEMISTRY. 139According to the views of Willstatter and Mallison, the reduc-tion of quercetin to allocyanidin takes the following course :HThe production of cyanidin itself from quercetin is of the greatestinterest from more than one point of view. I n the first place,since yuercetin itself can be synthesised, the reduction methodbridges the gap between the synthetic product and the naturallyoccurring colouring matter, and a synthesis of the latter has thusbeen accomplished. Again, this reaction clears up the question ofthe constitution of the colouring matter, and thus allows us todecide upon probable structural formulze for the kindred antho-cyanidins. It also helps us to trace the relation between the flavoneseries and the anthocyanins.The anthocyanins in their acid-freeforms are isomeric. with flavone derivatives which have one atomless of oxygen attached to their benzene nuclei. The flavones andflavonols contain the pyrylium nucleus in an oxidised condition.Thus the relation between members of the two groups is not closestin isomeric compounds, but, instead, we must compare similarlysubstituted compounds in pairs where the anthocyanin contains anextra atom of oxygen. For example, cyanidin is isomeric withfisetin, but i t is closely related t o quercetin, as we have seen.The following structures have been proposed as the: best repre-sentations of the various chlorides :OH ? 0Cyariidiii chloride.Pela rgo ti idi n chloride140 ANNUAL REPORTS ON THE PROGRESS OF CHEMISTRY.OHQ10HO CHDelphinidin chloride.OH 0 OH 9‘ F’I I IlC.OH\.. / I i i/C.O~~e\-/ \A/0& O A A /-\OH Hc)/\A--/-\OMe\A/ OHHO CH HO11 y rtillidiii chloride, Oenidin chloride.OH F’ 0Malvidiii chloride.Various yellow and white flowers have been examined, and fromthem colouring materials of the anthocyanin group have beenobtained by reduction methods.26 Apparently the reaction givesrise to no anthocyanidin, and if care be taken not to carry thereduction process too far, it is not necessary to reoxidise the product in order to obtain the pigment. This investigation has thrownlight upon certain botanical speculations with regard to the pro-duction of the anthocyanins from yellow colouring materials.Willstatter and Mieg’s method of extraction has been applied t oobtain the fusariurn pigment.27 From Fusarium orobanchus twodifferent colouring matters were produced. One of these is a yellowanthocyanin pigment, which is soluble in water or in alcohol a t90°, whilst the other is a red carrotene which resembles that ofWillstatter and Mieg in many respects, but differs from it inbeing more readily soluble in chloroform than in carbondisulphide.The carrotene from fusarium gives the usual d o u rreactions with acid and alkalis, contains no nitrogen, and crystal-lises in plates. Solventshave different effects on the colour of the carrotene, according totheir nature and the experimental conditions. For example, in26 A.E. Everest, Proc. Rmj. s’oc., 1914, [B], 87, 444; A . , i, 978,27 Bezssonoff, Compt. rend., 1914, 159, 448 ; A , , i, 1135.It has not yet been completely analysedORGANIC CHEMISTRY. 141cymene the substance forms a reddish-violet solution, but on raisingthe temperature to the boiling point the colour changes to clearyellow, and as the solution cools again the red tint slowly reappears.Alteration in tint on exposure to air is found in the case of benzeneo r chloroform solutions of the pigment., where an originally reddish-violet liquid becomes blue.The absorption spectra of the three varieties of the carrotene,violet (in boiling alcohol), reddish-violet (in benzene), and yellow(in cymene after boiling), have been examined, and appear to con-tain three absorption bands, which are shifted in position in pass-ing from the violet to the red compound.The yellow pigment which occurs along with the carrotenebehaves like a weak acid, and easily unites with bases.It appearsto be accompanied by sugar, but it has not yet been proved that itis a glucoside.The roots of Datisca cannabina contain varying quantities of twocolouring matters; one of these is methoxylated, whilst the secondcontains no methoxyl group, and has the composition Cl,Hlo06.By benzoylation and acetylation28 i t was shown that four of theoxygen atoms belong to hydroxyl groups, and from the generalbehaviour of the substance i t is deduced that it belongs t o theflavone class., being isomeric with fisetin and luteolin.When hydrolysed with alkali, datiscetin (as the pigment is called)yields salicylic acid, whereas when bromine is employed as acleavage agent both bromosalicylic acid and tribromophloroglucinolhave been identified among the products.These data suggest thatdatiscetiri is probably 1 : 3 : 1’-trihydroxyflavonol,0 OHso that i t is very closely allied to morin.The colouring matter of the blackberry (Rubus discolor) hasbeen examined, and several lakes have been prepared from the dye,but so far very little information with regard t o its constitutionhas been 0btained.~9Crocetin,30 the pigment of saffron, has been found to have thecomposition ClOHl,O2. Various salts of the compound have beenprepared and analysed, but in this case, also, our knowledge of thecomposition of the substance is in an elementary stage.28 J.Leskiewicz and L. Marchlewski, Ber., 1914, 47, 1599 ; A . , i, 856.29 G. Vecchi, Staz. sperim. agrar. ital., 1914, 47, 60 ; A . , i, 978.3O F. Decker, Arch. Phrm., 1914, 252, 139 ; A , , i, 979142 ANNUAL REPORTS ON THE PROGRESS OF CHEMISTRY.The dye lokao, or China green,31 obtained from Rhamuus chboro-plmra, appears to be much more complex than the' other pigmentswhich have recently attracted atbention. The ammonium salt(ammonium lokaonate) has the composition C4,,H4,0,,-NH4, and,when treated with oxalic acid, liberates lokaonic acid, C42H46025,a bluish-black substance with a metallic lustre. When tseated withacids, the ammonium salt decomposes into rhamnose and lokanicacid, C,,H,O,,.The last substance is evidently a phloroglucinolderivative. The following scheme shows some of the decompositionsof lokaonic acid and its derivatives:Lokaonic acidILokanic acid + rhamnoseC42H46025J. H2S0436'31 C6H1305HNO, I KOHJ.Delokanic acid + phloroglucinol +, NitrophloroglucinolC H 3*O*CllH504 C6H603 4 HNO,Oxalic acid + C,H,O,N =2-nitro-5-methoxybenzoic acid ?ChZorophyll.I n the course of last year, the study of the constitution of chloro-phyll advanced rapidly, and the main lines of its structure appearto have been determined with some accuracy. The present yearhas not seen any great advance in this direction, and most of thework which has been carried out deals with the properties ratherthan with the constitution of the substance.It will be recalled that Willstatter's results pointed t o the con-stancy of the chlorophyll quotient in plants; this view hasbeen contested by Borowska and Marchlewski,32 who state that thequotient varies, not only from species to species, but even in plantsof the same species which are subjected to different external con-ditions.Thus if the absorption spectrum of the pigment is ex-amined, i t is found t o vary in different plants belonging to thesame species, and it is affected even by the stage of growth towhich the plant has attained. According to these authors, theabsolute and relative amounts of the bluish-green neochlorophyllproduced in given cases depend on the character of the soil ina2 H.Borowska and L. Marchlewski, Biochem. ZeiLwh., 1913, 57, 423 ; A., i, 73.A. Riidiger, Arch. Phnrrn., 1914, 252, 165 ; A . , i, 979ORGANIC CHEMISTRY. 143which the plant is grown; then, in turn, the quantity of pigmentprment regulates the power of the plant t o absorb light of a suit-able wave-length, and this, finally, controls the synthetic processesby means of which the plant tissues are formed.Another paper of interest on this question has been publishedby Stoklasa, 6ebor and Senft.g3 According to them, phosphorusenters largely into the processes whereby chlorophyll is produced.They regard chlorophyll as being separable into three differenttypes of compound:(1) Phzeophorbin and its metallic derivatives.(2) Phaeophytin and the phaeophytides.(3) Chlorolecithins of phzeophorbinphosphatides.The third class are compounds of phzeophorbin or of phzeophorbinwith phosphorglycerol.The phosphoric acid is united to theglycerol esters of unsaturatad acids or hydroxy-acids. The formeracids are produced in the plants during the spring and summer,and are converted into the corresponding hydroxy-acids by oxida-tion. The change of colour of leaves in autumn depends upon thehydrolytic fission of chlorophyll and the formation of phzeophytinand phosphatides, which have a brownish colour, and do not maskthe yellow and red tints of the carotins and xanthophyll. Mag-nesium, calcium, and potassium are the main metallic componentsof the chlorophyll group, and thO first of these is supposed to bean active co-worker with the phosphorus in the growth and meta-bolism of plants.The decomposition of chlorophyll in the light and in the darkhas been the subject of extensive investigations.34 When the pig-ment is exposed to light, aldehydes are liberated, and a t the sametime a reagent makes its appearance which is capable of decom-posing potassium io'dide.As is to be expected, the action of lightof different wave-lengths is not t'he same, the red rays having moreinfluence than those of the blue end of the spectrum. Rays of allwave-lengths are active, however, for with a much longer exposureto blue light the same results are obtained as are observed with ashort exposure to red rays. The results are similar whether thechlorophyll is bleached in the plant fibre or is extracted and thenbleached.The presence of oxygen is necessary in these photo-decomposi-tions, and the action of the oxygen is evidently not catalytic, as itdisappears during the reaction.Hydrogen peroxide attacks chlorophyll both in the light and inthe dark.When light is excluded, the reaction is slower than33 J. Stoklasa, J. sebor, and E. Senft, Bot. Zentr., 30, i, 167; A , , i, 423.s4 H. Wager, Proc. Roy. S'oc., 1914, [B], 87, 386; A., i, 561144 ANNUAL REPORTS ON THE PROGRESS OF CHEMISTRY.when the specimen is illuminated. I n daylight the reaction doesnot appear t o be much more rapid than the decomposition of chloro-phyll in presence of oxygen under the influence of light.The results of these investigations suggest that the productionof sugars and starch in the green leaf may not be due t o the directphotosynthesis from carbon dioxide in water, but may be broughtabout by the photeoxidation of chlorophyll, with the productionof aldehyde and a subsequent polymerisation of the aldehyde intosugar.With regard to the actual mechanism of the photo-oxidationprocess, it seems probable that hydrogen peroxide plays a part inthe reaction.Formaldehyde se’ems to be a main product in thedegradation of chlorophyll.35 The resulh obtained in these twoseries of researches should be carefully collated.The pigments of brown a l p have been submitted to investiga-tion, and a long-debated question appears to have been settled.36Chlorophyll as such has been shown to exist in these plants; and,in addition, there are three yellow, non-nitrogenous pigments :fucoxanthin, carotin, and xanthophyll.The molecular proportiouof chlorophyll to the yellow pigments is higher in the land plantsthan in the a l g aPhyllocyanin and phylloxanthin have been submitted t o furtherinvestigation.37 The former substance was obtained by the actionof hydrochloric acid on chlorophyllan, and subsequent purificationof the product. On treatment with 1 per cent. aqueous potassiumhydroxide, phyllocyanin yields a mixture of anhydrophyllotaoninand some other substance. This paper also contains a criticism ofWillstatter’s researches on the action of alkalis on chlorophyll.Pyrrole Deriua tives.The interaction between the Grignard reagent and compounds ofthe pyrrole group appears still obscure.Two possible views of themechanism of the reaction have been put forward, one of whichpresupposes a direct action of the Grignard reagent on a carbonatom of the pyrrole nucleus, whilst the other hypothesis dependsupon the assumption that the reaction occurs in two stages, first,an attack upon the imino-group of the pyrrole ring, and then sub-sequent wandering of the *MgX group t o a point in the nucleus.An attempt has been made to settle the question by utilisingN-methylpyrrole as one of the reagents.38 Here there is noC. H. Warner, Proc. Roy. Soc., 1914, [B], 87, 378; A., i, 563.:M R. Willstatter and H. J. Page, Anualen, 1914, 4044, 237 ; A., i, 708.37 H.Malarski and L. Marchlewski, Biochem. Zeitsch., 1913, 57, 112 ; A . , i, 72.38 I<, Hess and F. Wissing, Ber., 1914, 4’7, 1416; A., i, 725ORGANIC CHEMISTRY. 145hydrogen atom attached to the nitrogen of the imino-group, so thatthe “two-stage” process may be left out of consideration. Theresults in this case are said to be similar to those obtained whenpyrroles containing the *NH* group are used. The work has beencriticised, however,39 on the ground that thP N-methylpyrrole usedwas not sufficiently pure, but contained an admixture of unsubsti-tuted pyrrole. It is to be hoped that the matter will be clearedup shortly, as the point is one of some interest.The alkylation of pyrroles has been closely studied recently, andthe matter is of importance on account of the relation between thepyrrole group and the colouring matter of the blood.When theGrignard reagent is allowed t o act on a secondary pyrrole, the firststage in the reaction appears to be the formation of an N-mag-nesium derivative, aad when this is treated with alkyl iodides theproduct is an alkylpyrrole. Thus, from methyl iodide and mag-nesium pyrryl bromide it is possible to obtain a mixture of alkyl-substituted pyrroles, of which the most plentiful is 3-methylpyrrole.The reaction appears to be a generalThe older method of alkylation, by the action of an alkyl iodideon sodium pyrrocarboxylate, has been further investigated in thecurrent year,41 as well as the direct action of alkyl iodides onp y r r o le s .42When chloroform is allowed to react with alkyl-substitutedpyrroles, the initial step in the process appears to be the elimina-tion of a molecule of hydrogen chloride with the formation of adichloromethylpyrrolenine.This reaction has been studied in thecase of 2 : 3 : 4 : 5-tetrachloropyrrole als0,43 and as it has beenfound to work with several other pyrrole derivatives, i t seems t obe a general one.Another general method applicable in the pyrrole series is theaction of sodamide on the dialkylallylacetophenones, the end-pro-ducts being pyrrole derivatives. Thus, when dimethylallylaceto-phenone is mixed with its own bulk of benzene and treated withone-fourth of the theoretical amount of sodamide, the chief pro-duct appears to be 2 : 4 : 4-trimethyl-5-pyrrolidone.44A third method leading to the synthesis of alkylated pyrroles,namely, the distillation of pyrrolecarboxylic acids, has been thesubject of extensive investigation,45 and numerous methyl deriv-atives of pyrrole have been produced in this manner.39 B.Oddo, Gaxxetta, 1914, 441, i, 706; A., i, 1142.40 B. Odd0 and R. Mameli, ibid., 1913, 43, ji, 504 ; A., i, 80.41 Ibid., 1914, 44, ii, 162; A., i, 1142.-la G. Plancher and 0. Ravenna, Atti R. Accad. Lincei, 1913, [v], 22, ii, 703 ;A., i, 319.44 A. Haller and E. Bauer, Compt. rend., 1914, 158, 1086 ; A., i, 724.4j H. Fischer and H. Rose, Zeitsch, physiol. Chem., 1914, 91, 184 ; A., i, 862.REP.-VOL. XI. L43 G. Plancher and T. Zambonini, ibicl., 712 ; A., i, 321146 ANNUAL REPORTS ON THE PROaRESS OF CHEMISTRY.It is well known that the direct action of halogens on pyrrolederivatives is so violent as to make it impossible t o prepare mono-halogen-substituted pyrroles in this manner.A simple method 46has now been devised by means of which these compounds caneasily be obtained. When magnesium methyl iodide is allowed toact on a secondary pyrrole, the hydrogen atom of the imino-groupis displaced by the radicle 0Mg-X (where X is a halogen atom).If this compound is treated with halogens a t low temperatures, areaction takes place in accordance wit'h the scheme:The monohalogen derivatives of pyrrole prepared according tothis method are unstable substances, and decompose even on keep-ing. Thus, bromopyrrole explodes violently when left in a sealedglass tube, the reaction apparently being represente,d by theequationDuring the last twelve months, the study of the indole grouphas not been pursued with the same vigour as was shown in theprevious year.The only point of interest that has arisen is thepreparation of 2 : 2I-di-indyl47 by heating together oxalo-o-toluidideand sodium amyloxide a t 360O.Further investigation of the picolide and pyrindole groups havebeen carried out, and an attempt has been made' t o determine theinfluence of substitution on the ease with which the compoundscan be prepared.48 Thus, when B-picoline, y-picoline, or pyridineitself is heated a t 200° with acetic anhydride, no picolide is pro-duced, alt,hough members of the picolide group are easily pro-duced from 2 : 4-dimethylpyridine and 6-phenyl-2-methylpyridineunder the same conditions.Pyridine as a Solvent and Catalyst.The use of pyridine as a solvent for inorganic and organicmaterials is not new, but some recent work49 suggests that itsapplication in this capacity requires circumspection before i t cansafely be regarded as a neutral solvent.When employed in the46 K. Hess and F. Wissing, Ber., 1914, 47, 1416 ; A., i, 725.47 U'. Madelung, D.R. -P. 262327 ; A., i, 89.48 M. Scholtz, Arch. Pharm., 1913, 251, 666 ; A., i, 431.49 M. Raffo and G. Rossi, Gnzxetta, 1914, 44, i, 104; A . , i, 572ORGANIC CHEMISTRY. 147case of sulphur compounds, i t appears to be attended with acerbain risk, as the following results show.Sulphur itself dissolves in pyridine on heating, but hydrogensulphide is liberated, pointing to a decomposition of the solvent.M7hen the solution is allowed to cool, a pitchy material is deposited,which has not yet been fully investigated.When organic sulphur derivatives are dissolved in pyridine, asimilar evolution of hydrogen sulphide takes place, but in thiscase the hydrogen is not furnished by the pyridine, but is derivedfrom the solute.The pyridine appears to act merely as a catalystfavouring the decomposition of the dissolved substance. Thus,thioacetamide dissolved in pyridine liberates hydrogen sulphide,and becomes converted into acetonitrile; thiobenzamide behavessimilarly, and produces benzonitrile.I n the case of diphenylthiocarbamide, the decomposition takesplace in two stages.The first phase of the reaction ends in theproduction of carbodiphenylimide,CS(NHPh), =H,S + C(:NPh),.As soon as this reaction has run t o its conclusion, a second reactionoccurs between the carbodiphenylimide and hydrogen sulphide,thus :SC(:NPh), + 3H2S = CS(NHPh), + NH,Ph +PhN:C(NHPh),+ C h .I n the presence of pyridine this reaction proceeds a t about 1 1 6 O(b. p. of pyridine), whereas in the absence of pyridine a tempera-ture 50° higher is necessary in order to bring i t about.Other analogous reactions have been examined, and i t appearsevident that in the case of sulphur compounds i t is necessary tobe extremely careful in using pyridine as a solvent on account ofthese peculiar effects which i t produces.D i ILapll t ha t h io xo P L ~ i~ ~n Su 7 I S .Reference may be m a ~ l s 5 ~ t o the fact that there appear to betwo series of salts ralated to dinaphthathioxin, and it seems notunlikely that they are somewhat akin to the twin series producedfrom compounds capable of assuming either the phenopyrylium orthe ordinary pyrone structare.The thioxonium salts are obtainedby the action of an acetyl haloid on naphthasulphonium-quinoneor by acting on the corresponding sulphoxide with perchloric orsulphuric acid in hot acetic acid solution. I n either of thesemanners, salts of the following type are formed:50 B. Ghosh and S. Smiles, T., 1914, 105, 1739.L 148 ANNUAL REPORTS ON THE PROGRESS OF CHEMISTRY.where X represents either a halogen atom or a univalent acidradicle.The true sulphoxide salts are colourless, whereas thethioxonium salts are highly coloured. The perchlorate has beenobtained by the use of cold reagents, and, on heating, i t is con-verted into the purple thioxonium perchlorate.The AIkaloids.The researches in this department of chemistry during the pre-sent year do not appear t o have thrown much light on the mainproblems of alkaloid constitution ; steady progress has been madein many directions, but there has been no example of the culmina-tion of an investigat'ion in the case of any of the more importantalkaloids. Instead, the researches of the year have been devotedt o that clearing up of minor problems which is the necessary pre-liminary to bro,ader outlooks.A careful comparison has been made between the properties ofquebrachine and those of that interesting alkaloid, yohimbine,51and the results seem to establish beyond doubt that the two sub-stances are identical.A discussion of the present state of our knowledge ofcantharidin52 leads to the conclusion that the formula of thissubstance is the following :CH*CO\ \CH2 CH I /\C H,/\-cH~co/0Cant haridin.Cantharidin behaves like an anhydride, being graduallyneutralised by alkali. When heated with an acetic acid solutionof hydrogen bromide, cantharidin gives rise to three products :(1) a dibromo-derivative, to which the structure I is ascribed;(2) a compound, C,,H,,0,Br2 ; and (3) bromohydrocantharic acid,which appears t o have the constitution represented by 11.Thelast substance can be resolved into its optically antipodic forms,and when the active forms are reconverted into cantharidin, thesynthetic cantharidin is found to be optically inactive, in51 E. Fourneau and H. J. Page, Bull. SOC. Phnrmacol., 1914, 21, 7 ; A.,52 J. Gadamer, Yerh. Ges. deut. Nnturforsch. Aertzte, 1913 (1914), 2, 494 ;i, 862.A . , i, 707ORGANIC CHEMISTRY. 149agreement with the symmetrical structure suggested in the formulaabove :\ CH( C0,H) \ PH"/ 7% FH/CH2\/CHBr*Co\CH,/\CHB~GO / \CH,/\CHR~--CO />o CH, CH >o YH2 FHCH, CH(1.) (II.1When heated with acetyl chloride, active cantharic acid yields afeebly active substance, isocantharidin, which is assumed to havethe following constitution :\ CH(OH)*CO PH"\/YH2 !?\c €3,,'\ c i3 ,-- co/'0 / ' CH, CReference may also be made to the preparation of some salinedouble compounds from cantharid~lethylenediamine.5~Met'anicotine 54 has been the subject of some investigationsduring the current year, but as the results do not lend themselvesto summarisation, the reader is referred t o the original paper.It has been shown that hydrastinine hydriodide55 can be p r epared by a very simple reaction from dihydrohydrastinine.Alcoholic solutions of iodine and dihydrohydrastinine are mixedin the presence of potassium acetate, and the required compoundis obtained from the solution in the usual manner.Various saltsof hydrastinine and its homologues 56 have been synthesised, whichhave structures represented by the following formula, in whichR = hydrogen, alkyl, aryl, or alkylaryl ; X = acidic group :CH,In order t o prepare them, N-acylalkyl derivatives of homo-piperony lamine,CH,:0,:C6H,* CH,-CH,*N<Aryl Acylare condense,d with phosphoric oxide, and the dihydroisoquinolinederivatives thus obtained are converted into the required salts.xi Farbwerlre vorm.Meister, Lucius & Briining, D.R.-P. 269661 ; A., i, 708.54 E. Maass and K. Zablinski, Ber., 1914, 47, 1164; A., i, 723.Farbenfabriken vorm. Friedr. Bayer & Co., D. R. -P. 267272 ; A., i, 79.58 H. Decker, D.R.-P. 267699; A., i, 198150 ANNUAL REPORTS ON THE PROGRESS OF CHEMISTRY.A synthesis of racemic hygrine has been carried out in thefollowing manner.57 Two isomeric 1-methylpyrrolidylpropanolswere prepared, one of which, on oxidation, should yield racemichygrine.It was found that the use of formaldehyde as a methyl-ating agent resulted in a simultaneous oxidation of the hydroxylgroup, and the products were obtained almost quantitatively.Comparison of the synthetic substances with the natural hygrineleads to the conclusion that the latter has the constitution repre-sented byThe oxidation of hydroxy-amines to amineketones, according t othe following scheme,:C*NHR . . . CH(0H) . . . +CH20=:C*NMeR . . . CO . . . + H20,appears t o be of general application.A considerable amount of work has been done in the tropinegroup, solanines 58 having been very carefully examined.Thehydrolysis of this compound produces solanidine-s, dextrose,d-galactose, and d-aldomethylpentose. There is t’hus only one atomof oxygen in the solanidine molecule, and it apparently belongsto a hydroxyl group. The nitrogen atom is present as part of animino-group in both solanine-s and solanidine-s.The action of nitrous acid on amines has been studied in thisseries of compounds, and the results obtained appear to justifythe following conclusions. The primary result of the action ofnitrous acid on a primary or secondary amine is the formation ofthe nitrite, NHRR’,HNO,. This compound then undergoes re-arrangement into an as-diliydroxyhydrazine derivative,NRR’*N(OH)2.The decomposition of this last substance may follow three liiies :(1) Anazoic decomposition.-Here nitrogen is evolved, and therelsults are as follows: ammonium nitrite yields water; primaryamines produce water and an alcohol; secondary amines of thetype NR, decompose, with the formation of a single alcohol, whilstsecondary unsymmetrical amines, NRR’, give two differentalcohols.(2) Dia.zoic decomposition.-Here ammonium nitriteyields nitrosoamine, NH,*NO, or isonitrosoimide, NH:NOH ;primary amines give either nitrosoamines, NHR-NO, or diazo-hydroxides, NR:N*OH, whilst secondary amines form truenitrosoccompounds, NRR’oNO. (3) Cyclazoic decomposition.-In57 K. Hess, Ber., 1913, 46, 4104 ; A., i, 199.58 G. Odd0 and M. Cesaris, Gazzetta, 1914, 44, ii, 181, 191, 209; A . , i, 1173,1174; F.Tutin and H. W. B. Clewer, T., 1014, 105, 559ORGANIC CHEMISTRY. 151this case, either stable azocyclic compounds or $-nitroso-compounds,such as I from aminocamphor nitrite(, or allonitroso-compounds likeI1 from +nitroso-oxindole, are obtained :(1.1 (11. )The following formula for morphine has been proposed byBraun,Sg but the evidence upon which it is based is of so lengthya nature that it cannot be summarised here, but must be soughtin the original paper:H2-H/ /\ / MeN >IX-H, \ I / ~~-\-/ \\ OH '\ / \/0I n the group of strychnine alkaloids, progress is being made inour knowledge of decomposition products, but during the presentyear there is no marked advance along general lines. Strychnine,GObrucine,61 acetylbrucinolone,62 and acetylstrychninolone 63 have allbeen oxidised by different methods, and the results are likely t oaid us later in settling the question of the structure of the variouscompounds, but i t is impossible to give a condensed account of thework in this place.The addition of bromine t o cinchotoxine has been investigated,but the results are of sixch a detailed character that no purposewould be served in giving a reproduction of them in this place.Here, also, the reader must be referred to the original.64A very complete study of the absorpt'ion spectra of variousalkaloids belonging t o the zsoquinoline group has been made,66 andi t has been shown that the presence of unreduced catechol nucleiin their molecules introduces a common factor into all the absorp-tion curves.Emetine and cephaeline also show a similar pecu-liarity, so that i t appears probable that their nuclei contain theJ. von Kraun, Ber., 1914, 47, 2312 ; A., i, 1138.H. Leuchs and H. Ranch, ibid., 3917 ; A., i, 199.6o H. Leuchs and G. Schwaebel, ibid., 1913, 46, 3693 ; A . , i, 79.62 Ibid., 1914, 47, 370 ; A . , i, 31i.63 H. Leuchs, ibid., 536 ; A., i, 429.64 G. Rohde and S. Meissner, ibid., 1507 ; A . , i, 71965 J. J. Dobbie and J. J. Fox, T., 1914, 105, 1639152 ANNUAL REPORTS ON THE PROGRESS OF CHEMISTRY.catechol group, a conclusion which is supported by chemicalevidence.The constitution of harmine and harmaline still remains in doubt.From the fact that harminic acid yields isonicotinic acid on oxida-tion, i t is concluded66 that the nitrogen of the pyrrole nucleuscannot be ortho- to the pyridine ring, as a nitropyridinecarboxylicacid might' have been expected in that case.Assuming this postu-late, which appears to be open to criticism, the following formulafor harmine has been suggested by 0. Fischer :The full evidence in favour of i t must, however, be looked for inthe original paper.The action of light on mixtures of various alkaloids and ketones67has led to some results of interest. When benzophenone is employedalong with coniine, sparteine, or piperine, benzopinacone is pro-duced, which is probably formed by the abstraction of hydrogenfrom the alkaloids. Nicotine condenses with benzophenone, andstrychnine appears t o be polymerised in the presence of the ketone.Narceine and bemophenone produce a reddish-brown, crystallinesubstance.Acetophenone with sparteine forins acetophenonepina-cone and a gelatinous base.The Purine Grozcp.The most important work in this branch of the subject is to befound in some papers by E. Fischer and his colleagues,68 which ap-parently mark the opening of another chapter in the history of thesynthesis of naturally occurring subst,ances. The objective of theinvestigations is the synthetic preparation of the nucleotides, butthe first' steps which have already been taken are concerned withthe preparation of glucosides of the purine derivatives. It is hopedthat the combination of these glucosides with phosphoric acid willsoon be carried out successfully, for this would add the nucleotidesto the other three classes, sugars, purines, and polypeptides, forour main knowledge of which we are indebted t o Fischer's labora-tory.Several of the purine glucosides have now been preparedtiti 0. Fischer, Ber., 1914, 47, 99 ; A., i, 316; compare Ann. Report, 1912, 159.67 E. Paternb, Gnxuctln, 1914, 44, ii, 99 ; A . , i, 1137.E. Fjscher and U. Helfeiich, Rcr., 1914, 47, 210 ; A . , i, 333 ; E. Fischer andI<. von Fodor, ibid., 1058 ; A , , i, 741ORGANIC CHEMISTRY. 153by the action of acetobromoglucose or its congeners on salts of thepurines with silver or other heavy metals. Considerable difficultyhas been found in purifying the products, as great care is necessaryt o prevent hydrolysis taking place.Glucosides of adenine, xan-thine, hypoxanthine, guanine, theophylline, and theobromine havebeen prepared, and in one or two cases galactosides and rhamno-sides have also been obtained.Various other papers dealing with the purine group have beenpublished, although they do not lend themselves to detaileddescription. Thus a series of studies in the degradation of theo-phylline and its allied compounds has appeared,69 some uramilderivatives have been examined,70 and Bhe salt-formation of barbi-turic acid and its analogues has been investigated.71The Colo.rcring Matter of the Blood.Last year an account was given of Willstiitter's views on theconstitution of haemin,72 and it appears desirable t o summarise herethe criticisms which Kiister 73 has passed upon them.Willstatter and Fischer regard the elimination of iron froinhEmin as a process which occurs in two stages, the first step beingthe addition of a halogen hydride to the hzemin molecule, and thenext process being the actual removal of iron to produce haemato-porphyrin.Kuster, in his latest publication, objects t o this sug-gestion, and thinks that t b matter is better expressed if theaddition of halogen hydride and the elimination of the iron atomare regarded as two loosely connected processes which are broughtabout through the agency of the same reagent.On Willstatter and Fischer's theory, the metallic atom in theliaemin molecule is united by means of two principal valencies t otwo pyrrole nitrogen atoms and by t'wo auxiliary valencies to twoother nitrogen atoms, each of which is a common member of apyrrole and a pyridine ring.The action of hydrobromic acidbreaks the latter (auxiliary) bonds, leaving the easily rupturedmain valencies intact. Kuster, on the other hand, brings forwardevidence that throws further light on the subject. H e finds thatwhen haematin is treated under pressure a t 130° with 10 per cent.hydrochloric acid, the iron is practically all removed from themolecule, whilst only traces of haematoporphyrin are produced ; b u twhen hzmin is similarly treated very little iron is displaced,69 H. Riltz and I<. Strufe, Aqtnalen, 1914, 404, 131, 137, 170 ; A . , i, 58G, 587,588.i" H. Biltz, {bid., 180, 199 ; A . , i, 589, 591.i1 lbid., 186 ; A ., i, 590.i3 W. Kiister, Zeilsch. physiol. Chwn., 1913, 88, 377 ; A . , i, 35.7z A m . Report, 1913, 157-8154 ANNUAL REPORTS ON THE PROGRESS OF CHEMISTRY.although much haematoporphorin is formed. Now it is knownt h a t the chlorine in hzemin cannot wholly be removed by the actionof alkali, which may be taken its a proof that the non-removablechlorine’ is attached to a quaternary nibrogen atom, and that thechlorine in hzemin is not united simply with iron or with nitrogen,but must be regarded as standing in some relation t o the twoelements jointly. This assumed division of the valency of thechlorine at’om is regarded by Kiister as the cause of the iron atombeing able t o unite with the basic nitrogen atoms so firmly as toforbid its removal except by the use of energetic reagents.Kuster urges in support of his view that if Willstiitter werecorrect and the chlorine atom played no great part in the matter,then it should be equally easy t o remove the iron atom fromchlorine-free derivatives of hzemin as i t is to eliminate the ironfrom hzemiii itself.This appears, however, to be contrary t o theexperimental evidence.Similar views are put forward with regard to the difference inreaction between hzemin and its methyl ester with respect t ohydrochloric acid ; the dimethyl ester, when thus treated, losesmuch more iron than does hxmin itself. Kuster regards this asa result of the relations subsisting between the carboxyl radiclesand the basic nitrogen atoms; the carboxyl and the ferrichloro-groups are supposed mutually t o saturate each other’s basicproperties, and when the dimethyl ester is substituted for thefree carboxyl compound, this activity is increased, so that in thedimethyl ester the bond between the chlorine atom and the basicnitrogen atom is weakened to such an extent that the main holdon the iron atom is maintained by the chlorine atom.Whence itfollows that the iron in this case becomes more readily separatedfrom the rest of the molecule.Further evidence is found in the behaviour of mesohzmin, whichKiister declares t o be lacking in the structural conditions postu-lated by Willstatter for the stability of its iron atom, although inactual practice the stability of hEmin and mesohzemin do notappear t o be very different.Kuster finally objects t o Willstatter’s rather ex cathedra state-ment that the existence of a sixteen-membered ring is improbable,and points out that in actual practice such a ring has beenobtained by R ~ g g l i .~ ~The paper must, however, be consulted in the original, as it doesnot lend itself to summarisation.A general review of the work which has been done on theoxidation of hzemin has been published by Hahn,75 but it is hardlyI?. Ruggli, Annalen, 1913, 399, 174 ; A , , 1913, i, 1106.75 A. Hahn, Zeitsch. Biol., 1914, 64, 141 ; A., i, 993ORGANIC CHEMISTRY. 155up to date, for in one section of the subject-the constituentpyrroles of hzmopyrrole oil-the most recent paper mentioned isdated August, 1912.Direct attempts to prepare ferropyrroles76 have been begun, butthe results, so far, have been rather in the nature of clearing theground. A t first it was decided t o investigate the reaction betweenthe Grignard reagent and ferric chloride, but the reaction did notseem t o promise much when the reagent was prepared fromaliphatic derivatives.Thus, magnesium ethyl bromide and ferricchloride yield ethyl chloride, ferrous chloride, and magnesiumchlorobromide in accordance with the equationMgEtBr + Fe,Cl, = EtC1+ 2FeC1, + MgBrC1.Magnesium phenyl bromide and magnesium benzyl bromideappeared to hold out better hopes, for, although the final productsin these cases were respectively diphenyl and dibenzyl, this resultpoints to the intermediate formation of an organo-ferrichloride,which then undergoes decompositiolz into the1 hydrocarbon, iron,and ferrous chloride, thus:2FePh2C1 = CBH,Ph + FeC1, + Fe.Success was attained in the preparation of di-2-methylindolyl f erri-chloride,which was obtained by the action of ferric chloride on a Grignardreagent derived from 2-methylindole.When the crude hzemopyrrole of Nencki and Zaleski is treatedwith diazonium salts, a portion of i t produces a red azo-dye, andthis particular constituent of the hzemopyrrole mixture has beentermed hzemopyrrole I.It has since been shown that this sub-stance is probably 2-methyl-5-ethylpyrrole, and in the present yearthe matter has been put beyond dispute by the synthesis of thelatter compound,77 which appears t o produce the same azo-dye asis obtained from the naturally occurring substance.The synthetic product is obtained by dry distillation of methyl-ethylmaleinimide with calcium hydroxide and zinc dust in a currentof carbon dioxide.When the end-product of this reaction istreated with ptoluenediazonium chloride, a mixture of two dyesis obtained, the structures of which are represented by I and 11,which concords with the results obtained with the natural hzemo-pyrrole (I):76 B. Oddo, Gazzettn, 1914, 44, ii, 268 ; A . , i, 1176.77 J. Grabowski and L. Marchlewski, Ber., 1914, 47, 2159; A, i, 993156 ANNUAL REPORTS ON THE PROGRESS OF CHEMISTRY,SMefEt gMe*EEtN H NHC,H:,Me*N,*C! C-C C*N,*C6H,Me\/ \/(1. ) ;C;iVe RE tC,H,Me*N,-C C*N,*C,H,Me\/NH(11.)The presence of these substances in hzmopyrrole is adduced asevidence that some earlier views will require revision.The isolation of the constituents of hEmopyrrole oil has beeiicarried out by a process of fractional precipitation with picric acid,and in this way i t has been found possible to isolate hzmopyrrole,phyllopyrrole, cryptopyrrole, and the '' haemopyrrole-a " of Pilotyand Stock.'*A separation by means of picrates has also been used in the casesof the esters of phono- and isophono-pyrrolecarboxylic acids.79 I nthis way it has been shown that the isophonopyrrolecarboxylic acidobtained from hEmine is identical with that derived from bilirubin.It is well known that when hydroxylamine is allowed to reactwith pyrrole, the products are ammonia and the dioxime of succin-aldehyde :Although this reaction is less easy to carry out in the case of thealkyl derivatives of pyrrole, it seemed likely that some interestingresults might be obtained in this way in the case of those alkyl-pyrroles which are connected with the colouring matter of theblood.Attempts have therefore been made in this direction.Tetramethylpyrrole, when treated with hydroxylamine, is con-verted into a dioxime of the formula C8H,,02:and examination shows that the parent ketone is the same com-pound as that which is obtained80 when methyl ethyl ketone issubmitted to the action of sunlight.Bilirubin and hamin were also treated with hydroxylamine inthe hope of producing a fission in their rings, but no results wereobtained. On the other hand, porphyrinogen undergoes a re-78 H. Fischer and K. Eismayer, Ber., 1914, 47, 1820 : A . , i, 886.79 H. Fischer and H. Rose, ibid., 791 ; A . , i, 429.80 H. Fisclier and W. Zimmermann, Zcitsch. physiol. Chcm., 1914, 89, 163 ;A., i, 318ORGANIC CHEMISTRY. 157action totally different from that which was anticipated, for,instead of a rupture of the ring, an oxidation reaction occurs,which yields as end-products mesoporphyrin and ammonia.An attempt has been made t o prepare coloured derivatives ofdipyrrylmethane with a view to a comparison between them and theblood pigment.81 Substances like the following have been pro-duced by condensing hamopyrrole-b and other compounds withchloroform :The absorption spectra of the compounds are somewhat similar tothat of bilirubin, but do not resemble the spectrum of the colour-ing matter of the blood.Similar attempts have been made, using perchloroethane, t o pro-duce the twin carbon bridge in compounds of the following type:AB>\C--C<kwhere the radicle " A " is N<"R'8R, and the radicle " G-CRB " isI 1 CR:$JR PU'II<F-,CR . The absorption spectra of the compounds derivedfrom hamopyrrole-b and perchloroethane resemble those of certainchlorophyll derivatives, such as phytochlorin-e. Other att'eniptshave been made, using chloral and glyoxal to form bridges, andthese also have given rise to products of some interest. Form-aldehyde was not found suitable.The Synthesis of Haematic Acid.The preparation of hamatic acid has proved to be comparativelysimple.82 Ethyl a-acetylglutarate was mixed with ether andpotassium cyaqide, and the calculated quantity of hydrochloricacid was then very gradually added, the reagents being constantlyshaken for two days. After separation from by-products, the resi-due was hydrolysed with 25 per cent. hydrochloric acid, extracbedwith ether, and the residue, after evaporation, was dissolved inwater and shaken with chloroform to remove further by-products.I n this way, a mixture of a tricarboxylic acid and a lactone wasproduced, which, on heating a t 183O under 9 mm. pressure, wasconverted into haematic acid. The process is outlined in thefollowing scheme :0. Piloty, J. Stock, and E. Dormaim, Ber., 1914, 47, 1124; A . , i, 755.s2 W. Kusterand J. Weller, ibid., 532 ; A., i, 442158 ANNUAL REPORTS ON THE PROGRESS OF CHEMISTRY.COMe*CH(C0,H)*CH2*CH2*C02H -+CMe(OH)(CO2H)*CH(C'O,H)*CH2*CH2*CO,H +~Me:$!*CH,*CH,*CO,Hco co0\/The Bile Pigment.83As both this subject and its nomenclature are somewhat com-plicated, i t may be well to recall the relations between certaincompounds. Haematin, which is the non-albuminous component ofthe blood pigment, forms the source of bilirubin, a constituent ofbile. The decomposition of bilirubin produces various substances,among which is bilirubic acid,84 and this, in its turn, can be de-graded to cryptopyrrole and isophonopyrrolecarboxylic acid.Cryptopyrrole has the structure I, whilst for isophonopyrrolecarb-oxylic acid two alternative formulae (I1 and 111) have beenproposed :fMe*sEt flUe*~-CH,*CH,*CO,H EMe*;CI1*CH2*CH,*CO,HCH CMe CH CMe CMe CHOr \/N H\/NH\/NH(1.1 (11.). (111.)When oxidised by another method, bilirubic acid yields differentproducts, methylethylmaleinimide (IV) and hzmatic acid (V) :$!Me:$!Et yMe:y*CH2*CH2* CO,Hco co co co\/NH\/N HFinally, the action of nitrous acid(IV), haematic acid (V), and theacid (VI), instead of the oximecarboxylic acid :The same oxime is(V. 1pro'duces methylethylmaleinimideoxime of phonopyrrolecarboxylicof the isomeric isophonopyrrole-~Me:$l*CH,*CH,*CO,HCO CNOH\/NH(VI. 1obtained by the action of nitrous acid ontrimethylpyrrolepropionic acid, which seems t o show that thepropionylcarboxylic acid radicle exerts considerable influence onthe structure of the product.EG For an account of previous work reference may be made t o Ann. Report, 1912,170.84 H. Pischer and H. Rose, Zeikch. physiol. Chem., 1914, 89, 255 ; A., i, 309ORGANIC CHEMISTRY. 159Now, since bilirubic acid, when treated with nitrous acid, doesnot produce an oxime of its own, but yields, instead, the oxime ofphonopyrrolecarboxylic acid (VI), the deduction may be drawnthat the “pyrrole acid ’’ in the bilirubic acid molecule is a tetra-substituted one, which leads t o the conclusion that the bilirubicacid structure is built up from a trimethylated pyrrolepropionicacid united with the basic pyrrole radicls through substitution ofone of the methyl hydrogen atoms by the basic nucleus. I n otherwords, bilirubic acid might have the formula V I I :fi Me-SEt Me*g*C H,*C H,*CO,H0H.C C-CH,-C CMe\/NH\/NH(VII. )Against this might, be urged that sodium ethoxide is almostwithout action on bilirubic acid, whereas substances containing twopyrrole nuclei unite(d by a methylene group are usually easilydecomposed by this reagent. It has now been proved, however,that bilirubic acid (and even hzmin) can be decomposed by meansof potassium methoxide, so that this difficulty vanishes, and theproposed formula for bilirubic acid appears t o be placed on asound basis.The constitution of bilirubin itself now lies open t o attack.The action of sodium ethoxide on this substance produces a yellowxanthobilirubic acid, which appears to stand in close relationshipwith bilirubic acid, since it yields the latter substance on reduc-tion. Its constitution seems t o be expressed by$!Me:$!Et ;cI! Me* E*CH,* CH,*CO,HCO U==CH-C CMe\/NH\/NHFurther considerations drawn from the behaviour of bilirubin andhemibilirubin, into which we cannot enter here, lead Fischer andRose to the following structural formula for bilirubin :CH,: C H *$--fi Me ; C i ’ M e * p 3 : CH,c c C C*OH\/\ /\/N H \ / N HNH / \ N HC = C/\/ \/\ C E 8 gM0 CMe*C*C H,* CH,*CO,H CO,H*CH,*CH,-C--CM160 ANNUAL REPORTS ON THE PROGRESS OF CHEMISTRY,and in consequence they consider that Willstatter’s suggestedformula for haemin is not above discussion.A comparison between the foregoing Report and those of previousyears which covered the same field will show that on the presentoccasion a greater number of subjects has been dealt with thanwas the case in 1912 or 1913. This has its advantages in givingan appearance of variety to the Report, but it has the correspond-ing drawback of leaving in the mind of the reporter a feeling thatperhaps he has not devoted sufficient space to certain importantfields. It is safe to say, however, that no Report is ever likelyto satisfy its author; his chief comfort is based on the hope thathis readers may not be so critical of shortcomings as he himselfis bound to be. Many subjects have been perforce omitted fromthe present Report simply because a proper description of themwould have entailed too much dwelling upon detail; other subjectshave been left out because any full discussion of them would havenecessitated somewhat lengthy historical introductions, for whichspace could not be spared, It must not, therefore, be supposedthat t,he foregoing sections include all the important work of theyear: rather should they be regarded as examples of the kind ofinvestigations which are being carried out in the various fieldsunder review.A. W. STEWART