ORGANIC CHEMISTRY.PART I.-ALIPHATIC DIVISION.To have been responsible for a section of the Annual Report fora period of seven years, during the greater part of which thenormal course of research was interrupted by the war, has involvedon occasions the exercise of a capacity to make the most of limitedmaterial, but the facility thus acquired has not enabled the writerto disguise the fact that, in the past year, very little progress hasbew made in the aliphatic series. The difficulties recentlyencountered in dealing with publications describing unfinished ordisjointed work have been more acute than ever, and, if the ex-perience of one research laboratory is an index of the combinedexperience of laboratories in general, it is easy t o account for thestationary position of this branch of the subject.Researches which were in progress in 1914 have been completedand published in the interval, or, rather, have in many cases beenpublished without being completed, and there has been but littleopportunity to finish any new investigations commenced duringthe past twelve months.No doubt this state of affairs is transi-tory, yet it is safe to predict that the future topia of researchdealt with in this section will differ widely from those which havebeen discussed in the past. Chemists have altered their per-spective in the course of five critical years, and many may find itdifficult to revive an interest in compounds and reactions whichare harmless or have no market value. I f such be the case, andif utilitarian research is in any sense to overshadow scholarlystudy, the prospect is deplorable, and the aliphatic series will beone of the first t.0 suffer.This tendency of the times is already apparent in severalbranches of the subject where there is little of scientific interestto record.Other factors have, of course, been operative. Poli-tical unrest has terminated most of the valuable work on aliphatichydrocarbons, and must also bear the responsibility for many othergaps in the Report. I n addition, the death of Emil Fiacher will556 ANNUAL REPORTS ON THE PROGRESS OF CHEMISTRY.profoundly affect further progress in more than one branch oforganic chemistry, and a heavy task thus devolves on those whostill have the opportunity and still retain the desire to unravelthe mysteries of carbon compounds.It is hoped that’the pages of future Reporb will show that thisresponsibility has been realised and successfully undertaken.Hydrocarbons.There appears to be every prospect that, for some time to come,most of the investigations on aliphatic hydrocarbons will fall withinthe province of the Reports on Applied Chemistry.The numberof the publications dealing with these compounds certainly remainsvery much as before, but, for evident reasons, the topics examinedhave become more and more technical in their nature, and con-sequently, on the present occasion, discussion may be limited to acomparatively small number of papers.The somewhat confusing literature on caoutchouc problems, withits contradictory results and periodical corrections, has practicallyvanished from.the pages of the journals. I n addition, theresearches on the complex unsaturated hydrocarbons, whichformerly appeared regularly from Russian sources, have been com-pletely suspended, and these facts are in themselves sufficient torestrict the present section of the Report both in scope and variety.On the other hand, two distinct types of research on hydrocarbonscan be readily recognised as engaging most attention. One is theutilisation of natural hydrocarbon gases as sources of aliphaticcompounds, and the other, it need scarcely be said, is concernedwith the reactions of acetylene. Although in each type potentialtechnical application has been the directing factor, yet the resultsobtained are frequently valuable from the purely theoretical pointof view.Considering the somewhat wide range of the problems dealt within these investigations, it is practically impossible to preserve asystematic arrangement of the subject-matter, but as the ‘‘ crack-ing” of natural paraffins has been the object of a considerableamount of research, this problem may be dealt with in the firstplace.A number of simple saturated hydrocarbons have beenselected as test substances, and a general scheme has been putforward in which the conversion of such compounds into aromatichydrocarbons is claimed to proceed through the consecutive form-ation of simple olefines, and, in turn, of higher olefines with con-jugated bonds.It is not unimportant to note that, in the par-ticular cases studied, the presence of metals does not favour thOBUANIC CHEMISTRY. 57production of cyclic hydrocarbons, and they may even act asnegative catalysts promoting degradation.1 Attention has alsobeen paid to the conditions under which tars are formed in thecourse of these reactions, and it is possible that research of thisdescription may in time throw light on the nature of the complextars formed by the pyrogenic decomposition of lignified celluloses.This would, of course, involve systematic research conducted onvery unpromising materials, but, a t the present time, there appearsto be a distinct tendency t o focus on the problems of polymerisation,and it is well that such is the case.More than passing referencemay thus be made to a paper2 in which the production of nitro-ethylene from 8-nitroethyl alcohol is described, as the product isnot only very easily polymerised, but it has been possible to dis-criminate between reagents which promote the change and thosewhich retard or even inhibit it. Water proved to be highlyeffective, but much inferior to alkalis, whilst, on the other hand,acids were found to be without action. I n this particular example,the polymerisation is not reversible, and the same holds true inthe case of acetylene when the change is promoted by the actionof the silent electric discharge. I n this connexion, details are nowavailable3 as t o the experimental methods by means of whichacetylene can be subjected to graded polymerisation by varyingthe temperature a t which the gas is exposed to the discharge, anda striking feature of the products obtained is their ready oxidationand high degree of unsaturation.Whatever the nature of thesepolymerides may be, they are evidently far removed in structurefrom benzenoid hydrocarbons, and it would appear that the con-version of acetylene into benzene is by no means so simple a reac-tion as is generally believed to be the case. The idea has evenbeen put forward that, in pyrogenic reactions a t all events, theformation of benzene from acetylene is a secondary change and ispreceded by profound decomposition involving the separation ofcarbon. This is opposed to much experience, but the experimentson which the claim is made were carefully selected and were appar-ently carried out under highly accurate conditions.4 It is alwayssurprising that pyrogenic reactions of acetylene give definiteresults, considering the somewhat rough and ready manner whichcharacterises much of the experimental work of this nature.Nevertheless, the list of compounds formed in this way continuesto grow, and has now been increased by the recognition of o-xylene,*l J.G. Davidson, J . Ind. Eng. Chem., 1918,10, 901 : A., i, 10.a H. Wieland and E. Sakellarios, Ber., 1919, 52, [B], 898 ; A., i, 307.H. P. Kaufmann, Annalen, 1918, 417, 34 ; A., i, 117.5. Hilpert, @a. Abhand. Kennt. Kohle, 1917, 1, 271 ; A., i, 38058 ANNUAL REPORTS ON THE PROGRESS OF CHEMISTRY.indene, and mesitylene as authentic products.Even the pyrogeniccondensation of acetylene and hydrogen sulphide, to which refer-ence has been made in previous Reports, has yielded fresh resultsin that a-thiotolen and thionaphthen have been detected asadditional products of the complex readions.5Research which is more attractive to the structural and syntheticchemist, and will doubtless yield valuable results in the future, isconcerned with the nature of the compounds formed when acetylenecombines with mercuric chloride. There is no necessity to discussthe importance of a subject which is evidently the key to the variedreactions of the hydrocarbon under the influence of catalysts, andit now seems reasonably certain that the white compound obtainedas the principal product when acetylene, acts on mercuric chloridein an aqueous system is, in reality, trichloromercuriacetaldehyde,(ClRg),C*CHO. This view is not without its critics, and it hasrecently been suggested,C on the basis of a somewhat remoteanalogy, that the substance is an additive compound of vinylalcohol.In any case, it is evident that water must have played apart in its formation, and that i t arises from a simpler inter-mediate compound. Twenty years ago it was observed that asecond additive compound possessing the composition C2H2,HgC12is formed in small amount during the reaction, and it is nowshown that, by using an alcoholic solution of the metallic chloride,it is possible to obtain the above compound comparatively rapidlyand in good yield.Not only so, but the substance is well definedand crystalline, and, considering the yields obtained and the experi-mental conditions employed, there can be no question but that itrepresents the first and simplest additive product. Good reasonsexist for allocating to it the structural formula ClHg*HC:CHCl,and in view of its convenient solubilities and potential reactivity, itmay be described as a synthetic reagent with a future.'Although the reactions of acetylene have now become wellstandardised, novel types are occasionally forthcoming, and anexample is furnished by the observat.ion that tetranitromethanecan be prepared by absorbing acetylene in concentrated nitric acidin the presence of a mercury salt. Quite apart from the value ofthe process as a method of preparation, it would appear that thefirst actioh of the acid is to form an unstable compound, which istransformed into the nitro-paraffin when heated with acid.8 Oneof the most gratifying signs in recent organic research is the desireR.Meyer and W. Meyer, Ber., 1918, 51, 1571 ; A., i, 72.W. Mmchot, Annalen, 1918, 417, 9 3 ; A., i, 145.K. J. P. Orton, Brit. Pat. 125000; A., i, 247.' D. L. Chapman and W. J. Jenkins, T., 1919, 115, 847ORGANIU UHEMISTRY. 59to detect, and if possible identify, every intermediate compoundformed in reactions, and this policy must make for progress in theend.Alcohols and their Derivatives.It seems advisable to discuss all types of alcohols under oneheading, as the numbereof papers dealing with this branch of thesubject is much smaller than usual.So far as simple monohydricalcohols are concerned, there is little to report, but it may be men-tioned that it is now possible t o identify 8-aminoethyl alcohol withcertainty through the agency of a number of derivatives, theproperties of which render them suitable as reference substances.Considering the importance which is attached to P-aminoethylalcohol, work of this description is by no means valueless, although,in the paper referred to: another objective can be discerned, asit is impossible to disguise the fact that the synthesis of anzestheticsbased on the novocaine model was the main goal of the work.Now that the preparation of butyl alcohol by the fermentationmekhod has been elevated into a manufacturing process, researchin the butane series will be much facilitated.Fermentation butylalcohol is, however, a mixture, but it is possible to isolate the purenormal form by taking advantage of the fact that the sodium saltof butyl salicylate is easily crystallised, and may thereafter bedecomposed with water. The purified m-butyl salicylate thusliberated is then converted into the alcohol in the usual way.10Another paper describing results which may find application in thelaboratory is concerned mainly with the preparation of amylenefrom commercial amyl alcohol by the catalytic action of aluminiumoxide. Although practical directions are now provided wherebyuniform yields may be obtained, this reaction has already beenstudied in some detail, and perhaps greater importance should beattached to the description of a method by means of which amylenemay be converted into tert.-amyl alcohol displaying a high degreeof purity.11A considerable amount of work has likewise been directed towardsimproving the preparation of ally1 alcohol and of unsaturatedalcohols generally, but the results thus obtained do not appear tocall for detailed reference.Before leaving the subject, however,mention should be made of a paper in which it is suggested thatthe formulae at present assigned to a number of important repre-S. Frlinkel and M. Cornelius, Ber., 1918, 51, 1654 ; A., i, 66.lo K, J. P. Orton and D. C. Jones, T., 1919, 115, 1194.l1 R.Ad-, 0. Kamm, and C. 8. Marvel, J . Amer. Chem. Soc., 1918, 40, -1950; A., i, 6160 ANNUAL REPORTS ON THE PROGRESS OF CHEMISTRY.sentatives of the doubly unsaturated alcohol series should be com-pletely rearranged. Basing his views on the evidence provided bya number of transformations undergone by a-citral, the author 12proposes to assign to geraniol the constitution previously allocatedto citronellol, and also to apply Tiemann’s formula for geraniol tonerol. This does not exhaust the list of modified formulae, as, ifthe above changes are accepted, the constitution of linalool mustlikewise be altered; but the validity of the argument offered insupport of these suggestions cannot well be discussed briefly,although i t should be mentioned that the relationship betweengeraniol and dipentene is readily explained in terms of the consti-tutions now proposed.The saturated glycols have also been the subject of considerablestudy both in the laboratory and in the factory, where, with theobject of producing materials which may replace glycerol for indus-trial purposes, much attention has been paid to the preparation ofmixed glycols from waste petroleums.The initial stage is, ofcourse, the cracking of the hydrocarbons to give mixtures of olefines,which are then saturated with chlorine and the products hydrolysedto the corresponding glycols. Numerous patents and papers havedealt with this. problem, and one may be quoted13 in which theclaim is made that the nitrates prepared from these mixed glycolsdisplay properties which render them in some respects superior t onitroglycerine as the basis of explmives. I n quite a different field,a number of individual glycols have been studied with regard tothe molecular transpositions which they undergo when dehydrated,and a lengthy series of papers deals with a large number of freshexamples of such changes.14 Most of the glycols chosen as testsubstances are highly substituted by phenyl groups, and muchingenuity has been displayed in utilising the Grignard reagent inthe preparation of compounds of diverse type. As a general rule,dehydration of these glycols by means of sulphuric acid gives riseto either a ketone or an aldehyde, and although it is difficult todecipher any important generalisation or novelties in the results,appreciative reference should be made to the work in view of thethorough fashion in which each result has been confirmed.It is only natural that researches on glycerol should reveal thedirecting influence of the technical importance of the compound,and it is gratifying t o note that many investigations conducted inthe factories display a very high standard.The technical pre-1s A. Verley, Bull. SOC. chim., 1919, (iv], 25, 68 ; A., i, 146.1s H. Hibbert, Met. and Chem. Eng., 1918, 19, 571 ; A., 1918, i, 521.l4 A. Orhkhoff, Bull. SOC. chim., 1919, [iv], 25, 108, and succeeding papera ;A,, i, 205ORUANIC CHEMISTRY. 61pbrhtion of the compound, more particularly by fermentationmethods, has recently been prominent, and although this subjectdoes not fall within the limits of the present section of the Report,it is interesting to note how, one by one, the methods are beingdisclosed whereby the Powers were enabled to supplement theirsupplies of glycerol during the War.Reference may, however, bemade to one factor which is doubtless of more than passing import-ance-the fact that the fermentation of sugar by means of aspecially resistant yeast is greatly affected by the presence of areducing agent, such as sodium sulphite. I n this way, the yield ofglycerol is much increased, a result which is in harmony with theview that glyceraldehyde and dihydroxyacetone are the essentialintermediate products in ordinary alcoholic f ermentation.15.16Before leaving the subject of glycerol, attention should bedirected to the application of spectroscopic methods to the problempresented by the existence of two crystalline modifications ofglyceryl trinitrate.17 It is now established that both forms giveidentical spectra in aqueous solution, and this disposes of the possi-bility of the two varieties representing chemical isomerides. Thisconclusion has been well supported by the spectrographic examin-ation of the complete series of partly nitrated glycerols, where, ofcourse, two isomerides are possible in each case, and do exist.Fresh progress has to be reported in the consecutive scheme ofinvestigations on optically active glycerol derivatives carried outby Abderhalden and his co-workers.Undaunted by many difficul-ties, they have continued their synthetical work with unflaggingzeal, and have now succeeded in obtaining the d- and E-varieties ofa glycerophosphoric acid, which has been isolated in the form oflithium salts displaying opposite activities.18 The method employedwas to introduce the phosphoric acid residue by acting on a pyridinesolution of d-a-bromohydrin by phosphoryl chloride.The productwas thereafter treated with water, and then followed a tediousseries of operations designed 60 remove halogens and pyridine andt o obtain crystalline glycerophosphates. There can be no doubt asto the success of the scheme, as lithium glycerophosphate wasultimately obtained in d- and Z-forms, the activities of which arein fair agreement, taking into account the tendency shown by suchcompounds t o form an epihydrin phosphate.The same authors19 have also undertaken the preparation of1 6 I(. Schweizer, Helv.Chim. Acta, 1919, 2, 167 ; A., i, 239.l* W. Connstein and K. Ludecke, Ber., 1919, 52, [B], 1385; A., i, 463.H. Hepworth, T., 1919,115, 840.E. Abderhalden and E. Eichwdd, Ber., 1918, 51, 1308 ; A., i, 3.lS Ibid., 1312 ; A., i, 262 ANNUAL REPORTS ON THE PROGRESS OF CHEMISTRY.active propylene glycol with the object of synthesising active fatsshowing a specific rotation sufficienily large to enable enzymehydrolysis to be studied polarimetrically with a fair degree ofaccuracy. As is frequently the case with compounds of simplemolecular magnitude, the experimental difficulties were consider-able, and although active propylenediamine was obtained, the com-pound proved ineffective as a source of the desired active glycol.An outline of the method which proved successful is given belowas an index of the many variations which are frequently necessaryin synthetical work of this description.Allylamine .?$ P-chloro-12-propylamine acid + d-/3-chloro-n-propylamine d-tartrate ?!?$ d-P-chloro-a-propanol -+ d-propyleneoxide _"c! d-propylene glycol.Unfortunately, the di-esters of the glycol do not appear to dis-play large rotations, but the work may nevertheless prove useful inthe future, and already an additional application of active proppleneoxide has to be recorded in that the compound has been used asthe indirect source of I-P-hydroxybutyric acid.A structuralscheme has thus been formulated to show the configuration of thisacid, and its relationship to Z-alanine, but, as in all such cases, thevalidity of the argument is dependent on the, unknown factor as towhether or not optical inversions take place during any of thereactions involved.As, on the present occasion, the number of publications on thechemistry of esters is too small to justify their discussion under aseparate heading, it may be well to include a t this stage referenceto a highly suggestive paper20 in which the hydrogenation of un-saturated fats is presented in a new light. That this importantchange, when promoted by the catalytic agency of metals, shouldshow a general resemblance to enzyme action was doubtless to beexpected, but the close analogy now displayed between hydrogen-ation and the enzymatic hydrolysis of glucosides is particularlystriking.The most significant result revealed by a study of thetime-absorption curves is that an unstable complex is formedbetween the catalysing metal and the unsaturated fat. This a tonce brings the reaction into line with standard cases of enzymehydrolysis, where similarly it has been shown that temporarycombination of the catalyst is an essential feature.137 ; A., ii, 403.KOH2O E. F. Armstrong and T. P. Hilditch, Proc. Roy. Soc., 1919, [A], 96ORGANIC CHEMISTRY. 63Aldehydes and Ketones.Although numerous papers dealing with the reactions ofaldehydes and ketones have appeared during the past year, practic-ally the whole of this work has been conducted on aromatic com-pounds, and although the results described are not without import-ance, it is impossible to discuss them in detail in this section.Accepting this restriction, there is but little left to report, withthe exception of isolated researches which involve some fundamentalpoints.F o r example, a return has been made to the problem ofclassifying ketones according to the relative reactivity of thecarbonyl group present, and reference may be made to two papersdescribing work of this nature. Thus, the condensation of a ketonewith ethyl cyanoacetate in the presence of an amino-compound isnot a general reaction, and may be either imperfect or completelyinhibited, according to the constitution of the ketone chosen.21This represents, of course, merely a rough and ready classificationof ketones in terms of one particular reaction, and is incapable ofexact treatment.On the other hand, it is possible to classifyketones according t o the quantitative reactivity of the carbonylgroup through the use of semicarbazide as a test reagent. I n thepaper now referred to,22 a distinction is properly drawn betweenreactivity and instability, and the idea is put forward that acarbonyl compound can react with a salt of semicarbazide only afterliberation of the free base. If this point be conceded, it followsthat the reactive power of any particular ketonic group may begauged either by using salts of sernicarbazide with acids of varyingstrength or by the addition of excess of an acid fa the correspond-ing salt until a point is reached a t which semicarbazone formationis inhibited owing to reverse action.The idea is ingenious, andhas been applied t o a series of aliphatic ketones, of which propylisopropyl ketone was found to be the most stable and methyl octylketone the most reactive. I n this case, also, the classification isbased on the behaviour towards one reagent, and the order inwhich the ketones are arranged in terms of increasing reactivitydoes not correspond with that deduced from their behaviour withphenylhydrazine, but this does not interfere with due appreciationof a somewhat unique paper and with the prospects opened out ofseparating closely related ketones by graded reaction.The only advance which has been noted in synthetical reactionsinvolving aldehydes is a new method for obtaining aedialdehydesor the corresponding keto-aldehydes.The process depends on the21 I. Guareschi, Gazzetta, 1918, 48, G , 83 ; A., i, 94.22 A. Michael, J. Amer. Chern. SOC., 1919,41, 393 8 A., i, 2 s 64 ANNUAL REPORTS ON THE PROGRESS OF CHEMISTRY.alkaline condensation of (a) a ketone or ( b ) an aldehyde with anethylenic aldehyde, according to the equationsCH,R*COR + CHR:CH*CHO = COR*CHR*CH;R*CH,*CHO,CH2R*CH0 + CHR:CH*CHO = CHO*CHR*CHR-CH,*CHO.It is true &at the examples illustrating the above changes havebeen selected entirely from t'he aromatic series, but the experi-mental details provided seem to indicate that purely aliphatic com-pounds may also be included within the scope of the general scheme.I n any case, the paper in question23 deserves mention, if only forthe light which it throws on many of the tangled resultsencountered in reactions conduct<ed on the diphenylethane series.The general problem of the conversion of aliphatic ketones intocyclic structures never seems to lose its attraction, but of severalpapers on this subject which have been published during the year,discussion may be limited to 0118.24S CN*CH,* CO CIA,,which has now been obtained in a purer condition than hit.herto,has been subjected to the action of simple reagents, and, as aresult, entirely new ideas are available as to the constitution ofcompounds previously regarded as thiazole derivatives.Thus, onsaturating thiocyanoacetone with hydrogen chloride, it is convertedinto 2-chloro-4-methylthiazole (11), a result which is consistentwith the view that the ketone reacts in one of the possible enolicformsThiocyanoacetonep - y\/--i, CIJ,*C CCI RH-5 CH;C cb H N N(1.) UJ.1It is, however, conceivable that the tautomeride (I) might undergorearrangement, involving transference of hydrogen to nitrogen,followed by the closing of the ring through oxygen. Should thisoccur, the fundamental structure a t once diverges from the thiazoleconstitution, and a series of derivatives should exist related to thefiH-? CH,-C C:NH .\/0The name " rhodim" is suggested for this class of substance, andit is now shown that three methylrhodims are readily formed, one2s H.Meemvein, J . pr. Chem., 1918, [ii], 97, 225 ; A,, i, 21.24 J. Tcherniac, T., 1919,115, 1071ORGANIC CHEMISTRY. 65of which is the compound previously regarded, on very insecureevidence, as hydroxymethylthiazole. This does not exhaust thecorrections introduced in a highly interesting paper, as, conharyt o Hantzsch’s statement, the action of ammonia on thiocyano-acetone does not give rise to aminomethylthiazole, although theexperiment has been repeated on a generous scale and underconditions favourable to the formation of such a qompound.Acids and their Derivatives.I n last year’s Report, mention was made of the method ofidentifying common acids through the agency of their p-nitrobenzylesters, and this useful type of work has now been extended by theapplication of o-bromoacetophenone as a reagent.The propertiesof the phenacyl esters thus obtained are, on the whole, more suit;able than those formed from p-nitrobenzyl bromide for the charac-terisation of aliphatic acids. It is now possible to select either ofthese reagents, the choice being determined by how far the meltingpoint of the expected product is removed from that shown by closelyrelated compounds.25Before considering more complex subjects, a general paper 26 onthe oxidation of organic compounds by means of silver oxide maybe noted. The subject-matter of the research is naturally as muchconcerned with alcohols as with acids, but a number of widegeneralisations are drawn which reveal the conditions favourableto the formation of carboxy-compounds under mild conditions.Thus, in an alkaline system, silver oxide functions as an oxidiserwhen any two of the groups*CH2-OH, :CH*OH, or *CO,Hare combined with:CH*OH, :CO, or :C(OH),.On the other hand, in neutral or acid systems, oxidation proceedswhen the secondary alcohol group is united to *CO,H, CH,., :CHz,or even H.It is of importance to observe that, in alkaline media,increase in the concentration of alkali affects only the speed of thereaction and not the nature of the products or the proportions inwhich they are formed. I n order to appreciate fully the signifi-cance of these results, it is necessary to compare them with thoseobtained by Nef in his monumental research on the oxidation ofcarbohydrates by means of Fehling’s solution, when it will be seenthat both inquiries have led to very much the same conclusions.86 J.B. Rather and E. E. Reid, J . Amer. Chem. SOC., 1919, 41, 75 ; A.,i, 167.86 R. Behrend and K. Dreyer, Annalen, 1918,416, 203 ; A., i, 64.REP.-VOL. XVI. 66 ANNUAL REPORTS ON THE PROGRESS OF CHEMISTRY.It must be a common experience to find that entirely new ideaspresent themselves when the constitution of a compound is referredto a structural model different from that usually adopted. Asugar, for example, almost ceases to be a sugar when it is describedas a derivative of tetrahydrofuran, but it is occasionally soundpolicy and a safeguard against impressions becoming stereotypedto regard a carbohydrate in this light.A somewhat similar case,which has the merit of simplicity, has recently been furnished, asit has been painted out27 that maleic anhydride bears the samerelationship to furan as benzoquinone to benzene. The analogy,although imperfect, is fairly well maintained, and extends to theproperty of the anhydride to give coloured solutions in certainsolvents. The effect of saturation or substitution, respectively, todiminish and intensify this property supports the view that a fairparallel exists between the unsaturated anhydride and quinone.The combined results are suggestive, and certainly worthy ofattention.The constitutional problems presented by the glutaconic acidsare by no means simple, but, as a result of the systematic researchesof Thorpe and his co-workers, it has been possible for some yearsto provide an adequate explanation of the isomerism displayed bythese compounds.Reference to the Annual Reports for 1912 and1913 will indicate the stage which had been reached when theseinvestigations were interrupted by the War, and it will be recalledthat the three isomeric types of a substituted glutaconic acid maybe represented by the formulaeRR*CO,H FR*CO,H fiR*CO,HCR CHR CRCO,H*bHR -LR-CO,H &H R-CO,Htrans-Labile form. Normal form. &-Labile form.These formulze are based on numerous reactions, but it is never-theless difficult to substantiate the structure shown above for thenormal type by the production of evidence that addition in the1:S-positions can take place.Even the reaction with bromine,which proceeds regularly with the labile isomerides, does not takeplace without molecular rearrangement when applied to a normalform. I-n the case of the corresponding esters, however, it ispossible to discriminate sharply between the normal and labiletypes, and, a t the same time, to confirm the structure assigned tothe normal form by taking advantage of the reaction with ethylsodiocyanoacetate. Hitherto, the opinion has been held that,whereas a labile ester reacts smoothly with this reagent, a normal2' P. Pfeiffer and T. Bottler, Ber., 1918, 51, 1819 ; A., i, 62ORGANIC CHEMISTRY. 67form is incapable of reaction except under conditions which pro-mote transformation into a labile isomeride.This view can nolonger be maintained, as it is now shown28 that the ethyl ester ofP-methylglutaconic acid does actually enter into condensation withethyl cyanoacetate to give a small yield of the 1 : 3-additive pro-duct. There can be no doubt that the addition did involve theterminal carbon atoms, as the product, when hydrolysed, was con-verted into y-methylbut,ane-aPG-tricarboxylic. acid, the constitutionof which was known.CH,*CO,Et C H,*CO,HCH*CO,EtI AHMe L H M ~ CHMe + CH,(UK)*CO,Et --+ I + I 6 H *CO,E~ IC'H C0,E t C'H*CO,HbH(CN)*CO,Et UH,-CO,HThis result, which supplies valuable confirmation of the structuralideas frequently expressed in this series of papers, was not obtainedwithout considerable experimental difficulty.To turn to another topic, i t may be mentioned that some featurwof general interest are presented in what is presumably the firststep in a survey of the properties of aliphatic compounds in whichthe principal carbon chain is highly substituted by shorter chains.For example, a reaction so well explored as the condensation of aketone with the ester of an a-iodo-acid does not always proceed insuch a manner as to produce a hydroxy-derivative, as it is nowfound that the corresponding olefinic compound may be formedsimultaneously. This has opened up a route to the synthesis ofaSyG-tetramethylhexoic acid, CH,*[CHMe],-CO,H, and the generalinvestigation has incidentally furnished Willstatter 29 with anumber of new lactones or anhydro-acids which may possibly findapplication in the varied problems studied by him, even i f theirimportance is not apparent a t the present time.With regard to the metallic salts of aliphatic acids, it is evidentthat the pyrogenic decomposition of these compounds still con-tinues t o be a favourite topic of research, but, of numerous papersdescribing such work, reference need be made to only 0118.3~ Anattempt has been made t o classify the metallic formates accordingto the temperature conditions under which they decompose t o giveformaldehyde, and also according t o their capacity to yield methylalcohol and acetone, which are the outstanding secondary productsof the reaction.The combined results are of importance in con-nexion with the preparation of aldehydes by the dry distillation2* J.F. Thorpe, T., 1919, 115, 679.29 R. Willstiitter and D. Hatt, Annalen, 1919, 418, 148 ; A., i, 431.80 K. A. Hofmann and H. Schibsted, Ber., 1918, 51, 1398; A., i, 7.0 68 ANNUAL REPORTS ON THE PROGRESS OF CHEMISTRY.method, as it has been shown in the course of the work that zincformate is the most suitable substance for the production of form-aldehyde.It is impossible to leave the consideration of metallic salts with-out reference to a type of investigation which, although moreappropriately dealt with in another part of the Report, has been inthe past the object of appreciative reference in this section. Inrecent years, our ideas on the nature of soap solutions have alteredprofoundly, largely owing to the discovery that such solutions dis-play an unexpectedly large conductivity in concentrations wherethe result cannot be attributed to hydrolysis.The explanationwhich has been put forward to account for this behaviour is thatan aggregation of charged ions forms the nucleus of the colloidalparticle, and that the system thus produced, termed the “ionicmicelle,” is responsible for a large proportion of the conductivity.The general terms of this theory have frequently been expressedin the, course of a lengthy series of investigations on soap solutions,and in the latest contribution31 it has been formulated in greaterdetail and supported by a series of conductivity, f reezing-point,and vapour pressure determinations carried out on solutions ofpure soaps a t ordinary temperatures. A study of the results, andmore particularly of the form of the conductivity curves, showsthat the theory accounts adequately for the known facts, and, inview of the increasing importance now attached to the electricalcondition of colloids, its elaboration is both important andopportune.Halogen Compounds.I n normal times, difficulty was always encountered by thereviewer in any attemph to discuss researches on halogen deriv-atives under one general heading, as the synthetic applications ofthese compounds penetrate into every branch of the subject.Thepast year has, however, been exceptional, as, in the aliphatic seriesa t least, the use of halogen compounds in syntheses has not onlybeen restricted in scope, but presents no feature of novelty.Onthe other hand, several papers have dealt with improved methodsof preparing simple haloids, and reference may be made to someexamples which may reasonably be expected to find application inthe laboratory.The general development under which petroleum gases are nowbeing used as the starting material in the preparation of fattycompounds of varied type has been extended to the formation ofJ. W. McBein, (Miss) M. E. Laing, and A. F. Titley, T., 1919, 115, 1279ORGANIC CHEMISTRY. 69simple aliphatic haloids. For example,32 a natural gas consistingprincipally of methane containing a small proportion of ethane,has been found t o undergo progressive substitution when, in admix-ture with chlorine, it is passed through a heated tube containinga suitable catalyst.It is not surprising to find that, in the searchfor a catalysing medium, anti-gas charcoal was tried and found tobe highly effective, but, of necessity, the process gives mixtures ofchloro-compounds, although it appears possible to grade the reac-tion so as to yield either carbon tetrachloride or chloroform as theessential product. So far, this method of chlorination has beenconducted with comparatively small quantities of material, but itis probable that it may be developed into large-scale working, and,i f so there are many outlets for the mixture of chlorides whichmay thus be obtained. It is, however, doubtful if the method cancompete with the parallel process in which petroleums are crackedto give olefines, and these are converted in turn into the corre-sponding dichlorides, as, in each case, the products are mixturesand will probably be used mainly as solvents.Another method of preparing aliphatic chlorides, the utility ofwhich is equally doubtful in view ‘of the fact that mixtures ofisomerides are formed, depends on the direct action of hydrogenchloride on an alcohol, the mixed gases being led over aluminiumoxide a t a moderate temperature.Scrutiny of the resultsobtained 33 shows that, under these conditions, the tendency of thealcohol to pass into the corresponding olefine cannot be altogetherexcluded, and thus the products, except in the simplest cases, arenot individual compounds, but contain secondary, and eventertiary forms in addition to the primary chloride. Of greaterimportance from the point of view of laboratory working is theaccount of an improved method for preparing alkyl iodides.34 Theprocess is really a modification of that recommended by Walker, asthe appropriate alcohol is boiled in contact with a mixture ofyellow and red phosphorus, the iodine being introduced by means ofthe refluxing liquid.The working details supplied in this latestcontribution to the scheme for facilitating the laboratory pre-paration of research reagents will be welcome, as it is a distinctadvantage t o have a safe and economical method of preparing thesimple iodides a t the rate of several kilograms a day.It is evident that, despite prolonged investigation, the complica-sa G.W. Jones and V. C. Allison, J . I d . Eng. Chem., 1919, 11, 639; A.,s3 P. Ssbatier and A. Mailhe, Compt. rend., 1919, 169, 122 ; A., i, 430.34 R. Adam and V. Voorhees, J . Amer. Chem. Xoc., 1919, 41, 789; A * ,i, 429.i, 30670 ANNUAL REPORTS ON THE PROGRESS OF CHEMISTRY.tions encountered in attempts to1 prepare chloroform by electrolyticmethods are far from being removed. Great possibilities are con-tained in such a process, but the usual practical difficulties regard-ing the choice of the most favourable conditions of current density,temperature, and concentration are increased by the fact that theaccumulation of alkali a t the cathode has the effect of decomposingthe liberated chloroform.By the adoption, however, of a neutral-isation electrode it is possible35 to minimisu this loss, but neverthe-less the method is complicated and somewhat uncertain in itsresu1t.s. Even under carefully standardised conditions the yieldsof chloroform obtained from acetone are throughout inferior t o thosegiven by alcohol when the electrolysis is conducted in the presenceof the chlorides of alkali o r alkaline-earth metals. This is not sur-prising in view of the fact that when alcohol is employed the firststep of the reaction is the formation of acetaldehyde, the conversionof which intol chlosoforrn by means of calcium hpochlorite isstated 36 to be instantaneous and quantitative.As already mentioned, no1 novel synthetical uses of halogen com-pounds have been noted, but a passing referenoe may be made tosome complicated results which have been obtained by the inter-action of magnesium phenyl bromide and halogenated ethanes.Theessential feature of these reactions is the fact that when substitutionof the paraffin molecule has been effected by different halogenatoms, these are in part eliminated under the action of the reagent.Although little uniformity can be discerned in these results, thOinrestigation will prove useful to those who have occasion to actwith Grignard reagents on pdy-halogen compounds.37 Greater satis-faction will be found in the study of an investigation 38 in which theaction of Grignard reagents on the esters olf aliphatic dib<asic acidshas been controlled, so that the change is limited to one carboxy-alkyl group.A considerable amount of attention has in the pastbeen paid to the1 problem of modifying the Grignard reaction, soas t o attack preferentially one of two groups which are apparentlysymmetrical. I n the case of di-esters, complete reaction on normallines should give a ditertiary glycol or a closely related compound,and practically without exception results of this nature have beenobtained when compounds of the type of diethyl oxalate, malonate,or succinate are subjected t o the action of Grignard reagents. Itis now shown, however, that the limitation of the reaction to oneposition is largely a matter olf restricting the proportion of mag-85 J. Feyer, Zeitsch. Elektrochem., 1919, 25, 115 : A., i, 305.38 S .Utheim, Brit. Pat. 116094; A., 1918, i, 521.3 7 F. Swarts, Bull. SOC. chim., 1919, [iv], 25, 145 ; A., i, 247.3 8 H. Hppwcrth, T., 1919,115, 1203ORGANIC CHEMISTRY. 71nesium alkyl haloid employed, although the special method adoptedto' incorporate the reacting substances may also be a factor. I n thisway it has been possible tol prepare a-hydroxy-a-ethylbutyric acidfrom ethyl oxalate and magnesium ethyl bromide, and this singleexample will be sufficient to illustrate the1 nature of the reaction.This appears to be perfectly general, as the only exception encoun-tered in a number of test cases was that of ethyl malonate, whereenolisation interfered with the normal coarse of the change.It is always of interest to encounter transformations from thealiphatic to the aromatic series which can be described as synthesesin the best sense of that expression, and another examples9 has beenadded to the existing 'list in that malonyl chloride reacts withacetone with the elimination of two molecular proportions of hydro-gen chloride, and the formation of phloroglucinol as the essentialproduct.The blackboard representation of 'this reaction will doubt-less, 'on account of its simplicity, find application in the lecture-room,and it is more than probable that an intermediate product formedin the condensation will prove useful in the laboratory, if the con-stitution assigned to1 it, namely, C'H,~CO*CH2*CO*CH,*COC1, iscorrect. A compound which has a six-carbon chain, is a diketone,and also an acid chloride is a t once stamped as embodying endlesspossibilities.Optical Activity.The idea was expressed i n last year's Report that it is inadvis-able to limit the1 review of optical activity entirely to examplesselected from the aliphatic series, and on the present occasion effectis given to1 the suggestion then put forward that all types of o'pti-cally active compounds should be discussed.It is difficult to1 presentthis subject in any order which has the merit of logical continuity,and, in such a case, it may be well to begin by an account of newresolutions which have been effected during the year. Of severalexamples, most interest will be attached to1 the revised specificrotation of tropic acid, the two active farms of which have beenisolated under conditions which guarantee their optical purityCuriously enough, this has been achieved in the course of twoseparate investigations which differed widely so far as the initialtopic of research was concerned.With regard t o one of theseinquiries it must be admitted that the detailed stereoohemistry ofthe alkaloids is beyond the scope of this section of the Report, buti t should nevertheless be stated that a distinct advance has beenmade in characterising two out of the eight possible activeT. Komninos, Compt. rend., 1918,167, 781 ;:A,, i, 672 ANNUAL REPORTS ON THE PROGRESS OF CHEMISTRY.hyoscines. Starting from the feebly active hydrobromidm obtainedin the manufacture of Z-hyoscine, the d-form has been isolatedthrough the agency of d-u-bromol-?r-camphorsulphonic acid. In addi-tion, ghyoscine was hydrolysed, not by alkali, as is usually the case,but by means of acid, and in this way Z-tropic acid was obtained,displaying an activity somewhat higher than the standard valuehitherto hccepted. It was thus desirable h repeat t h O resolutionof tropic acid so as to establish the maximum rotatory power, andthis was accomplished by the successive use of quinine and quin-idins.40 By a coincidence1 the values obtained by King were con-firmed in the course of an investigatioa,41 in which the standardsynthetical methods of preparing r-tropic acid have been tested andfound to be in many respects unsatisfactory.Exact working detailsare now provided, by means of which the colmpound may be obtainedi n good yield by the operation of a synthetical scheme in whichacetophenonecyanohydrin is the starting material.I n resolving theacid, quinine was used ta separate the d-form, and morphine wasfound to give the most satisfactory results in the isolation of theoptical isomeride. The rotatory powers od this important acid havethus been established with a high degree of accuracy.Turning to reactions conducted on optically active compounds,Abderhalden’s recent work on derivatives of glycerol has alreadybeen dealt with, so that it is psesible to pass t o the considerationof examples which illuslxate in holw far the mechanism o t reactionsmay be indicated through a study of optical changes.On firstinspection the subject of racemisation does not, appear t o be apromising source of new ideas regarding the prooess of ester hydro-lysis, but. nevertheless the! complexity of such changes is wellrevealed by considering the optical eff ecta encountered in hydrolys-ing ethyl Z-mandelate under different conditions .42 It may now beregarded as a general rule that alcoholic potassium hydroxide exer-cises a more powerful racemising elffeet than aqueous alkali in thehydrolysis of an active ester, and the result cannot be attributedto the action of the alkali on the liberated acid, as this is a, minoreffect compared with the direct racemisation of the non-hydrollysedester. It follows that the mechanism of hydrolysis is differentaccording as aqueous or alcoholic alkali is used, and it is reasonableto assume that potassium ethonide rather than potassium hydroxideis to be regarded as the essential racemising agent.It is now sug-gested that, in the case of aqueous hydroxide, an additive compoundis formed which afterwards loses the elements of alcohol, and thusH. King, T., 1919, 115, 476.A. McKenzie and J. K. Wood, ibid., 828.A. McKenzie and H. Wren, ibid., 602:; A., i, 326ORGANIC CHEMISTRY, 73gives the potassium salt directly as shown below in the specificexample of ethyl Lmandelate :OH OH OH. . . .____... 8(1.1 (11.) (I=)As the groups added and eliminated are not immediately connectedwith the asymmetric carbon atom, it follows that if (I) is lzvorota-tory, (111) will be active in the same sense. On the other hand,when alcoholic potassium hydroxide is employed it may be assumedthat the above compound (11) is replaced in part by (IT), so thatthe elements of alcohol can be eliminated only by the formation ofOH OH/\OK OEtw .1 w.1the unsaturated type (V). In the presence of water this would passinto the two active forms of mandelic wter, and as the5e would beproduced in equal amounts total racemisation would result. Thetheory is in harmony with the facts, and explains the partial race-misation of the non-hydrolysed ethyl Lmandelate which remainswhen the pure active ester is treated with alkali in insufficientamount t o cause complete hydrolysis. The combined results andthe use which has been made of them constitute a g o d example ofthe application of optical methds in tracing the mechanism ofreactions generally regarded as simple.In an entirely different field, optical changes have been used toexpand our ideas regarding the reactions of the halogenated suc-cinic acids.I n this work, Holmberg has frequently emphasised theidea that the removal of halogen from them compounds involvesthe transient formation of active malwlacbne,CO,H*CH *CH,*CO.The existence of such a lactone has been predicted largely from theresults (of physical measurements, but the compound has now beenisolated in r- and d-forms. This has been possible through thepreparation of pure iodoauccinic acid, the optical behaviour ofL - 0 1D74 ANNUAL REPORTS ON THE PROGRESS OF CHEMISTRY.which, it may b mentioned, is very similar t o that displayed bythe corresponding bromo-compound.*3 When treated with silveroxide and water 6-malo-lactone was farmed, and the properties ofthis product agree closely with those ascribed to it before its isola-tion had been acoomplished.Thus, when the lactone ring wasopened by means of acid some racemisation took place, but the malicacid produced contained excess of the Z-form. With alkali no race-misatioa occurred, but d-malic acid was alone prlduced.44 Thisresult must be regarded as highly satisfactory, as is also the factthat the views formerly expressed45 as to the reaction betweenZ-brcrmosuccinic acid and potassium xanthate have now been experi-mentally confirmed.46It is inevitable that papers which involve closely sustained andoontinuous argument cannot readily ble reviewed within the narrowcompass of the Reports, and this applies t o the latest contribution 47to the vexed question of the causes which are responsible for anoma-lous dispersion.Sttention is thus directed to the original paper,which deals with the rotation dispersion of the higher ethereal tar-trates, and introdurn a number of azguments opposed to the viewthat abnormal dispersion is to be attributed t o the co-existence ofdynamic isomerides displaying different rotatolry powers. It isequally difficult to do justice to1 another paper,** which deals witha new quantitative generalisation governing optical activity in thes u g a group.The number of acid amides related t o the sugarswhich have been obtained in a pure condition has recently beenincreased, and, as pointed out last year, the optical activity of suchcorr,poands is largely dependent on the configuration of the groupsattached t o the a-carbon atom. It is now possible, through theanalysis of the rotations shown by the acid amides from the C4 tothe C, series, to ascribe a quantitative value to the optical effectcontributed by the afly8-asymmetric systems of a sugar. This mayappear to1 be a bold claim, particularly as the generalisation dependson the exact application of the principle of superposition to sugars,but, at the same time, it is evident that the values now quoted willserve as a valuable guide in determining the constitution of partlysubstituted aldoses and in indicating which hydroxyl group has beensubstituted.B.Holmberg, Arkiu Kern. Min. Geol., 1917,6, No. 23,33 ; A., 1918,i, 623.44 B. Holmberg, Svensk. Kem. Tidskr., 1918, 30, 190, 215; A., i, 309.45 Ann. Report, 1917, 83.46 B. Holmberg and R. J. Lenctnder, Arkiv Kern. Min. Geol., 1917, 6, No.47 P. F. Frankland and F. H. Garner, T . , 1919,115, 636.‘r4 C. S. Hudson and S. Komatsu, J. Arner. Chem. SOC., 1919, 41, 1141 ;17, 26 ; A., 1918, i, 529.A., i, 524ORGANIC CHEMISTRY. 75The first step in what may prove1 to be an inquiry of considerablesignificance in biology is marked by the preparation of d- andLforms of simple dyes containing an asymmetric ~ystem.4~ The workhas not proceeded far, but evidence has already been obtained thatthem optical isomerides are selectively absorbed by wool, and theprospect is thus opened out that they may ultimately be used in thestaining of sections so as to reveal more completely the chemicalconstitution of tissues.This field of research has not been exploredby the chemist, and there is ample scope for future developmentsob great importance.Car b oh y dra t es .Any advances which may recently have been made in our know-ledge of the carbohydrates are largely discounted by t.he fact, towhich reference has already been made, that the great pioneer insugar synthesis died during the year. This is not the occasion, evenif space permitted, to make any attempt to1 pay fitting tribute tothe inspiration and instinctive genius which characterised EmilFischer’s best work o’n the’ sugar group, and it is perhaps sufficientto say that, master in more than one branch of organic chemistry,he was, above all, the master of sugar chemistry.He has left behindhim a record in carbohydrate research which many may imitate,but none can excel.In so far as publications reflect the methods of the man, a worthymodel is discernible in Fischer’s papers which well repay carefulstudy even by those who have no special interest in thO sugars. Hehad an extraordinary capacity to formulate schemes of researchwhich wem apparently disconnected, and then finally to marshalthe remits in order, sol that each fitted into its appointed place,making a complete story.It is appropriate that, on this occa+m, a departure should bemade from the customary order in which the carbohydrates havebeen dealt with in recent Annual Reports, and to consider in thefirst place the publications which mark the close of Fischer’s career.It will be recalled that, in 1912, he propounded the view that certaintannins may be regarded as fully esterified glucoses, in whichdigdloyl residues substitute the hydroxyl groups.Despite manyunexpected difficulties and distractions, he continued with character-istic courage lm test his ideas by means of syntheses, and from thetime he succeeded in devising a method f o r the preparation of thepenta-acetyl-m- and pdigalloyl chlorides, it was evident that hehad secured the reagents which would make succees possiblle.He49 C. W. Porter and C. T. Hirst, J. Arner. Chem. Soc., 1919, 41, 1264; A.,i, 558.Df 76 ANNUAL REPORTS ON THE PROGRESS OF CHEMISTRY.lived to see his hopes realised. By acting on glucm with either ofthe above acid chlorides, the colrresponding penta-(penta-acetyl-digalloy1)-glucoses were obtained, and, on further treatment withaqueous sodium hydroxide, the acetyl groups were removed, withthe formation of a penta-digalloyl-glucose :Exclusive of the obvious m- and pisomerim which may be contri-buted by the digalloyl residues, the capacity of the glucose com-ponent to exist in diverse modifications has t o be taken into account,SD that the structure formulated above may be represented byseveral compounds.Fischer found, however, that pnta-(m-digall-oyl)-&glucose resembles natural Chinese tannin closely, and thecorresponding a-compound likewise shows a general agreement,although displaying a different specific rotatioq. I n work of thisdescription, divergency in specific rotations ought not tot weighheavily in judging the success of a synthesis. Scrutiny of some ofFischer's experimental results, such as the effect of acetylation onboth eynthetic and natural tannins, reveals the delicacy of thesecomplexes and their ready tendency t o undergo intramolecularchanges. It would have been in the highest degree surprising ifthe specific roltations of natural and synthetic tannins had COST&sponded exactly, particularly as these values are determined oncolloidal systems. It may thus be taken that a tannin synthesis hasbeen effected, but considering that our views on the structure ofsugars are a t present in a state of flux, it must be admitted thatthe problem lof the constitution of tannin is far from having beensolved.Perusal of the two papers 601 51 dwcribing the resu1t.ereviewed above indicates sufficiently the difficulties encountered, andthe publications deserve careful study in view of the suggestivevariations introduced into ardinary working methods. Incidentally,iil the come of the work a number of minor points have been clearedup, and the patience with which these side-issues have been examinedcommands admiration. For example, the existence of a definiteglucogallin in Chinese rhubarb has always been regarded as indirectevidence in tiupport of the idea that the glucose formed by thehydrolysis of tannin is, not adventitious, but represents a specificcleavage product, yet numerous attempts t o synthesisel this appa-rently simple compound have given most conflicting results.Therecan be no longer any doubt as t o the nature of glucogallin, whichhas been synthwised by the interaction of acetobromogluoose and6o E. Fischer and M. Bergmann, Ber., 1918, 51, 1760; A., i, 87.61 Ibid., 1919, 52, [B], 829; A., i, 278ORGANIC CHEMISTJZY. 77silver triacetylgallate, followed by removal of the acetyl group.This proves that glucogallin is 1 -galloyl-P-glucose, and the com-pound is thus differentiated from p-glucosidogallic acid.52Fischer’s views as to the structure of a typical tannin have beensupported indirectly by the observation 53 that, in the presence ofbo’ric acid the conductivity of tannin isolated from the gall-nut isnotably increased.&!oreover, the exaltation observed is consistentwith the idea that, although there are twenty-five hydroxyl groupsin the molecule, these are distributed in such a manner that onlyten pairs are favourably situated f o r combinatioa with the acid.This is a striking result, but unfortunately it does not throw anylight on the stereochemical condition of the sugar residue, and thesame result would be given by a complex consisting of a monosubsti-tuted glucose with the necessary hydroxyl groups in a single side-chain.It is appropriate that Fischer’s work on tannin shouldbe continued by those with whom he was associated in his earlierinvestigations on this subject. Freudenberg’s latest contributions tothe general problem include an attempt t o identify the unknownhexose present in hamameli-tannin,54 and the isolation from chebulicacid of a new crystalline tannin which apparently possesses the com-paratively simp10 composition of a digalloylglucose. Before leavingthis subject it may be mentioned that, in his closing papers, Fischeradopted the numerical methlod of indicating the position of sub-stituents in sugar derivatives, and the fact may be used as anargument, where other arguments have failed, in favour of theadoption od this measure1 of relief t o the distracted workers 011related topics.It need scarcely be1 pointed out that Fischer’s synthesis appliest o one particular type of tannin only, and that the, nature olf othmclasses olf tannins still remains obscure.A conspicuous example ofthis is furnished in the case oif hemlock tannin. I n the course of anexhaustive experimental study, it has been shown that this complexcontains no sugar chain, and the authors of an important paper,55*confronted as they were with a mass of difficult experimental results,prudently refrain from expressing any opinion as to the detailedconstitution of the compound.52 E. Fischer and M. Bergmann, Ber., 1918, 51, 1804 ; A . , i, 89.J. Brjeseken and W. M. Deems, Proc. K. Akad. Wetensch. Amsterdam.1919, 21, 907; A., i, 412.54 K.Freudenberg, Ber., 1919, 52, [B], 177 ; A., i, 215.b6 R. J. Manning and M. Nierenstein, T., 1919, 115, 66278 ANNUAL REPORTS ON THE PROGRESS OF CHEMISTRY.Glucosides.From the tannins it is but a ztep t.0 the consideratio'n of theglucosides proper, and here also it is appropriata that Fischer'slast work should have been crowned with success. Reference wasmade in the Annual Report for 1917 (p. 79) to the variation heintro'duced into glucosidel synthesis, whereby nitrile-glucosides maybe obtained from hydroxy-esters, and, although the experimentaldifficulties encountered are probably more severe than is indicatedin the published papers, there seems no doabt that the methods arewidely applicable and extend to cases where an aliphatic nitrile iscoupled with the glucose residue.To1 take a case in point, aoeto-bromoglucose has been oondensed with ethyl a-hydroxyisobutyrate,and, on treating the product with ammonia, the acetyl groups arelost with the, ultimate production of hydroxybutyramide glucoside :-0 -__-_I ICH,(OAC)*CH(OAC)*CH~[CH*OA~]~*CHB~ + + (CH,),C( OH) *CO,Et1-0-(CHJ 2Q.OCH,(OH)*CH(OH)*CH*[CH*OH],*fHCO*NH,Finally, dehydration of the amide gives the nitrileglucoside, which,in the example given, proveid to1 be linamari11.5~ Although thescheme given above represents the essential steps, considerable varia-tion in procedure is evidently necessary in order t o obtain crystallineproducts, and when the conversion of the acid amide into the nitrileis effected by means of phosphoryl chloride, the product requiresre-amtylation. Obvioasly, the general reaction may be modified soas to produce a wide variety of synthetic glucosides of the cyanu-genetic type, and the inteJresting cases furnished by the glucosideof glycollonitrile 57 and the corresponding celloside 58 give promisethat the chemistry of amygdalin may no'w be attacked synthetically.It is not a severe criticism t o state that the remaining publica-tions on glucosides fall far short of those just reviewed.The heroicefforts of Kiliani to unravel the1 complications of the digitalis glumcsides have not been suspended, although positive results are few innumber, and no more than a passing reference need be made to theti6 E. Fischer and G.Anger, Sitzungsber. K. Akad. Wiss. Berlin, 1918, 203 ;A., 1918, i, 626.57 E. Fischer, Ber., 1919, 52, [B], 197 ; A., i, 256.68 E. Fischer and G. Anger, ibid., 854 ; A., i, 256ORGANICJ CHEMISTRY. 79latest contributions to the subject.59, go. In difficult work of thisdescription it is somewhat disappointing tol find that complicationsare needlessly introduced through the occasional choice of a reagentwhich appears in the higheat degree unsuitable, but the devotionof the investigator is most commendable. Other work which,although only remotely concerned with the sugar group, similarlyarouses feelings of sympathetic admiration is the first step in yetanother attempt to isolate an enzyme in a state 09 analytical purity.The research referred to opens u p a large number of importantquestions, and as the programme contemplated is ambitious, it iswell that i t is in competent hands,m but, as a side-issue which mayyet prove, tot be important, it may be mentioned that the peroxydaseof the horse-radish is associated with a nitrogenous glucoside theproperties of which, so far as they are described, point to a closerelationship with glucosamine.Two years ago attention was directed in the Reports to the, factthat, when silver salicylate reacts with acetobromoglucose, twoisomeric products are formed and the explanati'on then offered astm the underlying mechanism of the reaction has now been con-f i r m d .6 2 It has long been recognised that when the silver saltmethod of esterification is applied.. to a-hydroxy-acids an abnormalresult is obtained in that, to some extent, the alkyl group intro-daoed substitutes the hydroxy-position.The same holds true in thecase of a-amino-acids, and the most obvious interpretation is thatin silver salts of the type mentioned, the metallic atom is attached,not only to the carboxyl, but also t o the, hydroxy- o r amino-poei-tion. According to this view, silver salicylate would be representedasand consequently the glucosides obtained from such a structure bythe actio'n of acetolbromolglucose would display the isomerism shownbelow :/-\*O-Glucose residueand \-/ /-\OH \-/CO-O-Glucose residue CO,HIt is a curious fact that, as a reagent, acetobrolmoglucose is speciallywell adapted for the display of this isomerism, of which several new59 H.Kiliani, Ber., 1918, 51, 1613 ; A., i, 90.6O Ibid., 1919, 52, [B], 200; A., i, 214.61 R. WillstZlitter and A. Stoll, AnmZen, 1918, 416, 21 ; A., 1918, i, 556.6* P. Karrer, C. Niigeli, and R. Weidmann, HeZv. Chim. Acla, 1919, 2,242 ; A., i, 33880 ANNUAL REPORTS ON THE PROQRESS OF CHEMISTRY.examples are now described. As an interesting side-issue of theresearch in question, it may be remarked that the resolution ofinactive mandelic acid was achieved by taking advantage of thedifferent solubilities of the glucmides, but evidently, if this methodis to receive application, due attention must be paid to the forma-tion, on the line6 indicated above, of glucosido-acids.Monosaccharides and their Dem’uatives.Tol turn to the apparently simple subject of the monosaccharidesthere is little of outetanding interest to1 report.As has recentlybeen the case, a number of improved methods of preparation ofsugarsl and related compounds have been described, and amongstthese may be mentioned convenient processes for the isolation ofrhamnae,63 and of gulonolwtone.64 There is generally somethingdefinite about methods of preparatioln, but it is difficult to expressan opinion on the new sugar f l d o s e , which, although claimed tobe an unknown aldohexme, displays a suspicious similarity t o galac-t 0 ~ e . 6 5 It is just possible that the compound may be a variety ofgalactose corresponding with one of the abnormal galactose penta-acetates, but an opinion on this point must in the meantime bewithheld.On fir& inspection, a research on the properties of y-hydroxy-valeraldehyde may seem to have little connexion with the sugargroup, but tthe compound, which has been obtained in an ingenionsmanner,66 is suitable for testing the idea that aldehydic andhydroxyl groups when separated by a chain of three carbon atomsundergo mutual rearrangement to give a butylene oxide.The alde-hyde in question certainly displays most of the reactions of a reduc-ing sugar, and as it gives a ‘‘ methylglucoside,” an additional argu-ment is thus forthcoming in support of the current view of sugarstructure. The critic may find some inconsistencies in the evidence,and it is unfortunate that the: absence of optical activity placeslimitations on the search f o r analogies, but this does not detractfrom the appreciation of the treatment of an interesting subject.Work od a similar nature might with advantage be conducted ona-, &, and 8-hydroxy-aldehydes in vielw of the obscure isomerismof sugars t o which attention has been directed in recent years.I naddition, the physical examination of sugars requires considerable63 E. P . Clark, J . Biol. Chem., 1919,38, 255 ; A., i, 387.64 F. B. La Forge, ibid., 1918,36, 347 ; A., i, 65.6s E. Takahashi, J . Toby0 Chem. Soc., 1919,4Q, 157 ; A., i, 387.e6 B. Helferich, Bet-., 1919,52, [B], 1123; A., i, 386ORGANIC CHEMISTRY. 81expansion, and thus a further study G7 of the mutarotation of glucoeeand fructose is welcome, even admitting that this phenomenon hasalready beeri thoroughly examined.The fact that temperatureaffects the rotation equilibrium of the ketose is not new, but it ishighly significant that, in this special case, mutarotation cannot beregarded as due simply to a change in the position of one hydroxylgroup. This adds to the evidence that fructoee exists in solution inmore than two forms.It is with considerable reluctance that no detailed account isgiven here of Levene's efforts t o advance the difficult chemistry ofthe amino-sugars, but the experimental treatment of the subject issuch that results leading to final conclusions as to configuration are,of necessity, delayed until the full scheme od research is complete.It would thus be premature, and also an injustice t o a carefullyconceived series of researches, t o discuss the intermediate results sofar contributed, but reference should bel madel to the papm whichhave appeared during the year.I n order that the investigationsmay be followed, it should be mentioned that the goal of the workis the allocation, to a definite configuration, of the amino-grouppresent in compounds of the glumsamine t p . The method ofattack is to study the corresponding amic acids, rather than thesugars to which they are rdated, and t o pay particular attentionto' the compounds which form epimeric pairs. The difficulty in pre-paring these hexoeamic acids or their epimerid-68 and the compli-cations involved in the removal of the aminoLgroup by nitrous acid:9can be fully appreciated only by thoae whcm work has led theminto this field.I n the latest publication on this subject 70 a distinctadvance has been made in that the epimeride of glucosamine hasbeen isolated and described. The reactions of the compound arenormal save in two respects. NO monocarboxylic acid has yet beenobtained from it, and, curiously enough, when treated with nitrousacid, no molecular dehydration took place, but saccharic acid wasformed. Nevertheless, the sugar is readily converted into the ana-logue of chitose, and this was isolated in the crystalline condition.Some confusion of ideas seems to exist a5 t o the reasons underlyingthe failure of epichitosamine to display mutarotation, and until thefree sugar has been obtained in u- and P-modificatioins in whichdefinite configurations can be allocated to the reducing groups, it isaltogether premature to delete from consideration a betaine struc-ture for this and analogous compounds.i, 256.67 J.M. Nelson and F. M. Beegle, J . Amer. Chem. SOC., 1919, 41, 659 ; A,,68 P. A. Levene, J . Bid. Chem., 1918,86, 73 ; A., 1918, i, 530.G s Ibid., 89; A,, 1918, i, 532. 'O Ibid., 1919, 39, 69; A., i, 47582 ANNUAL REPORTS ON THE PROGRESS OF CHEMISTRY.Disac charides.Recently the contributions under this heading have shown a dis-tinct falling away in number, but the quality of the work describedhas been well maintained. Synthesis by means of enzyme actioncontinues to make progress, and amongst new results may be men-tioned the formation of gentiobiose as one of the products obtainedin a reaction designed to produce glycol glucdsides by the syntheticaction of emulsin.71 I n view of the earlier synthesis of gentiobiosethe result is not surprising, but the auto-condensation of gluoose isnot restricted to' on0 type of coupling, oellobiose having been isolatedas the fourth definite product 6f the synthesis.72 The latter resultassumes a new importance when taken in conjunction with the factthat the chemistry of cellulose is now being attacked by means ofenzyme degradation, so! that the p i t i o n has been reached in which,by oppositely directed enzyme actions, the ascent from glucose andthe dement from cellulose are being studied.If, and where, suchlines of work meet is an interesting speculatioln.The value attached t o enzyme synthesis of disaccharides is great,and its importance in throwing light on the stereochemical condi-tion of the constituent h-exoses is well rmgnised, but, in the mean-time, the inner structure of the disaccharides must be elucidatedby methods more familiar to the organic chemist. The structuralstudy obf the disaccharides which depends on methylation, followedby identification of the hydrolysis products, has already given im-proved formule for sucrose and lactose. Results equally diagnostichave now been obtained with maltose. The fact that in this casethe sequence of reactions outlined above gives tetramethyl glucoseas one product has now been confirmed73 under conditions whichadmit of no dubiety of interpretation, but the research was compli-cated by an unexpected degradation encountered in the preparationof methylmaltoside, and thus the structure of the reducing glucosecomponent i n maltose remained undecided.This, however, has beenaccomplished by varying the experimental procedure.74 Startingfrom the free sugar, methylmaltoside was produced by the carerullyregulated action of methyl sulphate and sodium hydroxide. Thereaftar, the same reagents were effective in completing the alkylation,so that the final product was a heptamethyl methylmaltolside. Onhydrolysis, the substituted hexoses isolated proved to be tetramethylglucose and the form of trimethyl glucose which has been obtained11 E.Bourquelot and M. Bride], Compt. rend. !919,168, 263 ; A,, i, 137.'* J. C. Irvine andwJ. S. Dick, T., 1919, 115, 693. '' W. N. Haworth and (Miss) G. C. Leitch, {bid., 809.Ibid., 1016 ; A., i, 361ORBANI0 UHEMISTRY. 83from methylgluooside. Identification of these methylated sugarsaffords an experimental verification of Fischer's formula for maltose,which is evidently in good agreement with all the properties of thesugar.Polysaccharides.Whilst research on cellulose and its derivatives continue8 to beexceedingly active, nothing of a very definite nature has been noteddealing with the primary constitution of the most important of allpolysaccharides. Our ideas regarding thg chemical constitution ofcellulose cannot be dissociated from the physical nature of the com-plex and, a t the prment time, it is well to preserve an open mind onthe general question.Various reviews of past work have been con-tributed during the year by well-known investigators in this field,but these need not be discussed, and reference may be confined toone experimental result which, although nolt directly connected withcellulose, has a certain significance. It will be remembered thatPictet obtained Z-glucosan by the dry distillation of cellulose orstarch under diminished pressure, and that he put forward the ideathat these polysaccharides arise from the polymerisation of this parti-cular anhydroglucose along different lines. It is now claimed 75 that,in part, this expectation has been verified as, under the influence ofplatinum black, giucosan is transformed into an amorphous com-pound, (C6H1005)4, displaying the properties of a dextrin and inwhich optical activity is retained. 'The further expansion of thissubject will be watched with great interest,.and it may be noted inpassing that the chemistry of starch is at present attracting numerousworkers.Although in general the results obtained are beyond thesoope of this section of the Report, brief reference may be madet o the vigorous discussion which has centred round the claim thatformaldehyde effects a diastatic degradation of starch. The discus-sion has, in fact, expanded out of all proportion to1 the inherentvalue of the loriginal experimental evidence, but the controversynow appear3 t o be ended.It has been shown, for example;76 thatthe failure of starch to give the iodine reaction after treatment withformaldehyde cannot be accepted as valid evidence of degradation.It has even been found that unchanged starch may bfe recoveredquantitatively after treatment with formaldehyde.77 This is damag-ing evidence, and the idea that the formation of a loose additivecompound of formaldehyde and starch would account for all the75 A. Picfet, Hdv. C h h . Acta, 1918, I, 226; A., 1918, i, 527.v6 M. Jacoby, Ber., 1919, 52, [B], 658; A., i, 311.77 W. von Kaufmann and A. Lewite, iW., 616 : A., i, 31284 ANNUAL REPORTS ON THE PROURElSS OF CHEMISTRY.result& described by Woker is supporbd from other quarter~.~8* 79 Atthe same time, bio-chemists seem reluctant to abandon the idea thata parallel may be drawn between the action of diastase and that offormaldehyde, but the arguments produced are not convincing, andthe subject may be regarded as clwed.Nitrogen Compo.u72ds.So long aa discussion is restricted to substances which are essen-tially open chains, the publications (of the past year on aliphaticnitrogen compounds have been less numerous and less complicatedthan usual.Many of the papers describe new or improved methodsof preparing common reagents, and these may be considered in thefirst place. As is but natural, much attention has been given tothe question of utilising, in a profitable manner, the large quantitiesof organic shell-fillings which have recently been aocumulated, and,in the case of chloropicrin, it has been shown80 that the compoundcan be economically employed for the production of methylamine.When reduced by means of iron and hydrochloric acid, exceedinglygood yields of the amine salt are obtained, and, as is usually thecase in this particular type of reduction, it is possible to restrict theam'ount of acid to1 about 3 per cent.of the theoretical quantity. I nview of the applications of methylamine it is an important p i n tthat, under favourable conditions, only a small proportion ofammonium chloride is furmed, and it would appear that the con-centration of acid used is an essential factor in controlling thecourse of the reduction. An expIanation of thia result is found inanother research,81 where it is shown that chlloropicrin is graduallyresolved a t the boiling point into carbonyl chloride and nitrosylchloride :CCl,*NO, --+ COCl, + NOCl.It would thus appear that the formation of methylamine by reduc-tion is due to reaction of chloropicrin as such, whereas ammonia isto be regarded as derived from the decomposition products.It is perhaps doubtful if the element of danger involved in thepreparation and use of chloropicrin will permit of the applicationaf such a method on the large scale.It may be remarked that thereaction between ammonium chloride and formaldehyde, which wasrecommended by Werner as a convenient source vf methylamine,78 H. Sallinger, Ber., 1919, 52, [B], 651 ; A., i, 263.7O J. Wohlgemuth, Biochem.Zeibch., 1919, M, 213 ; A,, i, 361.80 P. F. Frankland, F. Challenger, and N. A. Nicholls, T., 1919, 115, 159.J. A. Gardner and F. W. Fox, iM., 1188ORGANIC CHEMISTRY. 86;gives g o d yields of the base, and Dome modifications of the workingdetails have been contributed in the course of the year.82Much ingenuity has also been expended on the development ofprocesses for separating primary, secondary, and tertiary amines,but the references for the most part are in the patent literature. Asan example, it may be stated that when a mixture of amines istreated with ethyl chloroformate, the tertiary form remains un-changed while the primary and secondary bases are regenerated byhydrolysis.83 Carbonyl chloride may also be employed tol separatesecondary and tertiary amines,a and in addition considerable sue-cess has attended the attempt to develop a practical method of wp-rating amines by partial neutraIisation with hydrochloric acid.85Before leaving the subject of amines mention should be made of asubstantial improvement in the method for preparing diawton-amine,86 in which the action of ammonia on acetone is greatly facili-tated by the addition of calcium chloride.This variation consti-tutes a distinct advantage, and its adoption not only gives enhancedyields, but reduces the recolvery of unaltered acetone and alcohmolto a minimum.Passing to a related subject, a solmewhat unexpected property ofacetobromoamide is described by Wohl,87 who has shown that thecompound can function as a brominating agent.The reactionseems to be applicable to' a large variety osf cases, and the conversionof phenol into y-bromophenol may be quoted as a sufficiently strik-ing example of its elfficacy. With regard to the mechanism of thechange, it would appear that a direct interchange of hydrogen andbromine occurs between the two reacting molecules, and, if thisproves to be the case, the prospect is opened out of conductingbrominations without the formation of hydrogen bromide as aninevitable and disturbing by-product. Obviously, this would inmaay cases be a highly desirable condition, particularly whenunsaturated or optically active compounds are being manipulated.It may be remarked that, for the time being, research on opticallyactive amino-compounds is in a state of suspension, but there is adistinct revival in the study of general synthetical reactions withoutreference t o stereochemical problems.Most of these investigationshave been oonducted on standard lines, and are extensions oif former82 H. I. Jones and R. Wheatley, J . Amer. Chem. SOC., 1918, 40, 1411 ; A.,1918, i, 527.a8 W. Rintoul, J. Thomas, and Nobel's Explosives Co., Ltd., Brit. Pat.127740 ; A,, i, 388.84 Ibid., 128372 ; A,, i, 433.85 E. A. Werner, T., 1919,115, 1010.87 A. Wohl, Ber., 1919, 52, [B], 51 ; A., i, 198.88 A. E. Everest, ibid., 58886 ANNUAL REPORTS ON TEIE PROGRESS OF CHEMISTRY.work, so that a t present de'tailed reference is unnecessary, as nop i n k of fundamental theoretical importance appear to be involved.Structural questions, such as the distribution of the nitrogen valen-ciea and the isomerism of the quaternary ammonium salts, still con-tinue t o attract workers, but here also is a lack of novelty, althoughmention may be made in passing of the abnormal salts isolated byMredekind in his studies of compounds clontaining two) asymmetricnitrogen atoms of unlike asymmetry. On the other hand, distinctprogress has bmeen made in the constitutional problems presented bythe carbamides, and some of the outstanding results are now dis-cuss&.It is frequently the case that intimate study of a reaction, gener-ally regarded as simple, reveals many unexpected complications, andmuch of the recent research on urea furniihee examples of this.I nparticular, the customary method of expressing the formatiotn ofcarbamide from carbonyl chloride, can no longer be claimed to give afaithful representation of what occurs, as the reaction appears tobe based oa the union of ammonia and cyanic acid in the keto-imino-f orm.89 I n addition, the latest contributions to the' study ofcarbamides not only lend strong support to the views expressedby Werner as to the constitutional changes undergone by these com-pounds, but. also1 describe new working methods which are of value.I n the case of a monosubstituted urea, evidence has been accumu-lated to show that the equilibrium:NH,*CO*NHR OH*C(:NH)*NHRis determined by the electrochemical chaxacter of the group R andthe co-existence of these forms is oonfirmed, in the particularexample of monomethylurea, by a quantitative study of the reactionwith nitrous acid.This can be dissected into two parts, in thesecond of which nitrosomethylurea is suddenly formed 90 af&r theevolution of nitrogen has ceased, and the subsequent conve:sion ofthe nitroso-colmpound into diazomethane by the agency of sodiumethoxide has disclosed the fact that alcohol may be employed as asolvent in the latter reaction. Contrary t o expectation, diazo-methane reacts with alcohol extremely slotwly, i f a t all, and it isthus possible to effect methylation of a hydroxy-compound by dis-solving the latter in an alcoholic solution of nitrosomethylurea andthen adding the calculated quantity ,of sodium alkyloxide.This is adistinct advance, on the customary method of dissolving, oh suspend-ing, a compound in an ethereal solution olf the methylating reagent,88 E. Wedekind and T. Goost, Ber., 1919, 52, [B], 446; A., i, 286.89 E. A. Werner, T., 1918, 113, 694.Ibid., 1919,115, 1093ORGANIC CHEMISTRY. 87and advantage has been taken of this useful variation to verify thestatement that urea is not affected by diazomethane.91 Even whendissolved in alcohol and kept in contact with the reagent for manyhours the urea r'emained unchanged, and this result, which is inhasmony with the view that urea is t o be regarded as possmsingthe cyclic structure in neutral solution, is in sharp contrast to thatobtained with thiourea under parallel conditions.In this ease, amethyl group becomes attached to sulphur with the formation of thehomologue, SMe*C(:NH)*NH,, thus adding to the evidence thatthiourea exists in neutral solution as an equilibrated mixture of :A survey of the literature 'on the use of diazomethane reveals asurprising number of cases where this modified procedure wouldhave proved an advantage, and it is well to direct attention to themethod.I n drawing up this Report the original papers have been con-sulted wherever possible, but before these came to hand copiousnotes had in many cases been prepared from the abstracts publishedby the Society. The writer is thus in a position to appreciate theunfailing accuracy of the synopses and, as on a previoas occasion,cannot close the Report without exprwing his indebtedness to theabetractors.JAMES COLQUHOUN IRVINE.PART II.-HOMOCYCLIC DIVISION.2% eoretical.IT is extremely difficult to determine the origin and authorship ofcertain theories in modern organio chemistry, and a reporter mayperhaps best note the trend of speculation and emphasise any pointson which there seems to be an almost general agreement.Duringthe past year several papers have appeared dealing inter alia withthe problem of orientation, and the connecting thread is un-doubtedly the theory that atoms and groups direct the positionin the molecule taken up by entering substituents or, alternatively,determine the constitution of an additive product as the result ofa certain effect on alternate atoms in a chain.H. J. Prim1 dis-O1 E. A. Werner, T., 1919,115, 1168.1 Chem. Weelcblad, 1918,15, 671 ; A., i, 7188 ANNUAL REPORTS ON THE PROGRESS OF CHEMISTRY.tinguishea between link-energy and atom-energy and the mannerin which the alternating effect is assumed to occur may begathered from the following quotationz: “The entry of anysubstituent X into the benzene ring must cause a change inthe relation between atom-energy and link-energy, both inthe substituent and in the nucleus. Two cases may arise.In the first, in which the link-energy between X and C,, themrbm atom to which X becomes attached, is greater than thatbetween C, and the hydrogen atom displaced; the atom-energy ofC, is therefore reduced, and to restore this as far as possible, thelink-energy between C, and its neighbours, C, and C3, is reduced,with the consequence that the link-energy between C, and C, andbetween C3 and C, is increased (C, and C, b&g the neighboursof C2 and C, remote from C,), and that between C, and C, andC, and C6 is diminished; C,, therefore, by the diminution of itslink-energy, receives an increase of atom-energy, and is thereforerendered more reactive.The effect of introducing X, therefore, isto make the para-carbon atom more reactive. In the second case,in which the link-energy between C, and the substituent is lessthan between C, and hydrogen, the redistribution of energy oitaesan increase in the atom energy of C, and C,, that is, of the carbonatoms in the meta-position.”D. Vorliinder 3 employs a very similar conception, that of variableinternal molecular strain, but uses + and - signs to illustrate thealternating effects of atoms and groups.He definitely states, how-ever, that his views are not based on valency theories, The condi-tion of strain supposed to exist in nitrobenzene (I) and aniline (11)is illustrated below, and it is assumed that reachions will occur in+- - ++a INH,+H ), H+ \, + \/such a manner as to diminish the link tensions so that the m-posi-tion will be attacked in nitrobenzene and the o- and p-positions inaniline. The whole theory resembles very much those of Flur-scheim and of Thiele. The alternate + - labelling of atoms in itPrim, b c . cit. Ber., 1919, 52, [B], 263 ; A., i, 319ORGANIC CHEMISTRY. 89- - -straight chain starting with those of recognised polarity (Cl, Br, I,0, N, S, Na, H) has also been applied by A.Lapworth 4 a4 a meansof explaining, and in certain cases of predicting, the direction ofchemical changes, such as substitution and addition. Lapworthsviews have an electrical basis, and his symbols indioate all the anionsand cations which the molecule could conceivably yield. Actual ion-isation is neither assumed nor excluded. I f as the result of thismethod of expression an unsaturated centre bears the + sign it willattract the - portion of the molecule added, and vice versa. Also,if a + atom acquires additional + character it becomes more reac-tive, and similarly - atoms enter into reactions more readily iftheir - character is enhanced as the result of the influence of thepolar atom from which the labelling commences.Exactly similarresults are obtained by the application of the present writer’s 5 viewson conjugation of partial valencies, primary and secondary. Forpractical purposes Lapworth‘s notation is the most convenient, anda few examples may be cited.+ + - - -+-I- - - ++ + Additioit of HRr to Propylene, H3C-GH===CH, + - + - + - + - + - + -The effect of all the hydrogen atoms is here carried through tothe unsaturated carbon atoms, and it is seen that the central atomis oveIwhelmingly positive, and the result of the reaction is accord-ingly the produotion of &opropyl bromide.Addition of Hydrogen Bromide to Vinyl Bromide andA ZZyl Byomide.- + -CH,=CHBrH--Br + -- + - +CH2=CH-C!H2BrRr--H- t-Addition of Hydrogen Cymide t o Unsaturated Ketones.+ - -CH=CH-&O-CN--B - +Brit.ASSOC., Bournemouth meeting, Sect, B. Ibid90 ANNUAL REPORTS ON THE PROGRESS OF CHEMISTRY.Substitution in the Aromat,ic Series.- + -OH+H\ /\ /H+\/ + vNO,If H-I n this connexion mention may be made of a most interestingobservation by W. H. Gough and J. F. Thorpe,6 who find thatalthough 0- and p-xylylene dibromides react with potassium cyanidein alccholic solution with the formation of xylylene dioyanides andthe intermediate bromo-cyanides cannot be isolated, it is an easymatter to obtain o-bromo-m-tolylacetonitrile (I) by the interactionof m-xylylene dibromide and potassium cyanide.I n accordance withthe + - rule, the three xylylene dibromides and intermediatebromucyanides would be formulated as shown below, and it is atonce evident that, whilst the introduction of the first cyan+groupenhances the reactivity of the remaining bromine atom in the 0- andpseries, i t diminishes it in the m-compound:f- +f- CH,&- I + -/\/ - \-\ -k /\/A\/F H2Brf-/ /\ \/ /CH,Br fI tl- \ - / I -I+ I It/vI I- i - CH,Br + -I- + CH2Brf -* The arrows: indicate:this cme the nitrogen.,CH,*CN C H , ~ N//\\/ I /\ / + \\ - /\/I +I-\/ I :I I\ + /I CH,Br - +UH,Brf - (1.1the atom from which the labelling commences, inThis work, taken a t random to illustrate the application of a6 T., 1919,115, 1155ORGANIC CHEMISTRY.91theory, is in itself a matter of considerable importance, and furtherdevelopments of our knowledge of half-stage reactions in symmetri-cal ccrnpounds will be welcomed. Most synthetical chemists havehad sad experiences of poor yields obtained in such processes.Molecular Rearrangement.L. Claiseni has continued the investigation of the remarkabletransformation o l substituted phenyl allyl ethers into allylphenols.The exhaustive allylation of phenol is effected in accordance withthe following scheme :No trace of 4-allylphenol or 2 : 4-diallylphenol is produced in thetransformation of phenyl allyl ether and o-allylphenyl allyl etherrespectively. 2 : 4-Diallylphenol is, however, obtained by eliminationof the carboxyl group from the product of complete allylation ofsalicylic acid.A similar device has resulted in a synthesis ofeugenol (IV). Methyl guaiacolcarboxylate (I) is converted into itaallyl ether (11) by boiling with allyl bromide, potassium carbonate,and a little potassium iodide in methyl ethyl ketone solution. TheOH O*C3H, OHM~O/)CO,M~ MeOf",CO,Me MeO/\lCO,H(1.1 (11.1 (In.)Me01 JCH,*CH:CH2(N.1\/ L C3H5 LHO/\\/' L. Claisen, 0. Eisleb, and F. Kremers, Annulen, 1919,418, 69 ; A., i, 26692 ANNUAL REPORTS ON THE PROGRESS OF CHEMISTRY.acid obtained on hydrolysis of the methyl ester yields o-eugenol andcarbon dioxide on baing heated, but the ester itself is very readilytransformed a t 230--240° into the methyl ester of 6-hydroxy-5-methoxy-3-allglbenzoic acid (111).The latter yields eugenol whentreated with dimethylaniline a t 1 60° :Pinncone-Pinacolin Transformation.The dehydrating action of zinc chloride converts l-methyl-1 -a-hydroxyethylcyclopentane (I) into 1 : 2-dimethyl-A~-cycZohexene(11), and the reaction is one of the smoothest known enlargementsof the cyclopentane ring 8 :CH,A considerable number of ccglycols containing aryl groups havebeen synthesised,g and their dehydration products examined.Normal results were observed.The Action of Benzilic Acid on Arylthiocarbimides.Becker and Bistrzycki 10 found that the addition of benzilic acidto phenylthiocarbimide did not yield the expected O-ester,NHPh*CS*O*CPh*COzH, but instead N-phenyl-S-benzhydrylthio-carbam a,tea- car box ylic Theyassumed at the time bhat the O-ester was the first product, and thatthis changed over to the substance actually obtained, *CS*O* becom-ing *CO*S*.In the case of benzilic acid the intermediate productcould not be isolated, but the assumption made has now been justi-fied by the study 11 of the addition of benzhydrol to phenylthiocarb-imide. The reaction is carried out in xylene solution with thesodium derivative of benzhydrol, and results in the formation of0 -benzhydryl N-phenylthiocarbamate, . NHPh*CS*O*CHPh,. Thetransformation to the S-ester, NHPh*CO*S*CHP&, may be accom-plished by boiling with acetic acid or by heating a t 130-135O or bya u d , NHP he C 0 S CP hz C 0,H.H. Meerwein, Annalen, 1918,417, 255 ; A., i, 162.A.Or6khoffYBuZl. SOC. chim., 1919, [iv], 25, 108, 111, 115, 174, 179, 182,186; A., i, 205, 206, 271, 272.lo Ber., 1914, 47, 3149 ; A,, 1914, i, 245.l1 A. Bettschart and A. Bistrzycki, HeZv. Chim. Acta, 1919, 2, 118; A , ,i, 207ORGANIC CHEMISTRY. 93cold hydrochloric acid. The change by acids is regarded as beingdue to hydrolysis with the formation of benzhydrol, which in theform of an ester adds on to the *CS* group. The intermediate stepmay then be written:/S*CHPb2 ...NTIPh*C/ -0,Ac ,‘0- CHPh,\ ..and the reaction is completed as indicated by the dotted line, benz-hydryl acetate being eliminated. This hypothesis is strongly sup-ported by the observation that the transformation may be effectedby heating the O-ester with a little benzhydryl acetate or bromide intoluene.NHPh-CS-O*CH,Ph,is stable towards boiling glacial acetic acid.The reaction between chlorotriphenylmethane and diarylaminesproceeds normally only in the case of p-tetramethyldiaminodi-phenylamine.In other cases investigated 12 there is molecularrearrangement,, and, for example, chlorotriphenylmethane anddiphenylamine yield p-anilinotetraphenylmethane (11). !The normalproduct (I) is obtained from tetraphenylhydrazine and triphenyl-methyl. It is converted into p-anilinotetraphenylmethane by heatingwith diphenylamine hydrochloride in benzene solution :CPb ,*NPh, CPh3*C,H,*NHPh(1.1 (11.)It is interesting that the benzyl derivative,New Reactions.A simple synthesis of phloroglucinol has been described.lSMalonyl chloride and acetone in the presence of calcium carbonateyield phloroglucinol and diacetoacetyl chloride,CH,*CO* CE€2*CO*CH2.COc1,which can be changed into the trihydric phenol by boiling water inthe presence of calcium carbonate.P-Resorcylaldehyde and 2 : 4 : 6-trihydroxybenzaldehyde have beenobtained 14 by an application of Hoesch’s synthesis. Hydrogenchloride is passed into an ethereal solution of resorcinol (or phloro-glucinol) and cyanogen bromide in the presence of zinc chloride.H. Wieland, B.Dolgow, and T. J. Albert, Ber., 1919, 52, [B], 893 ; A . ,i, 324.l3 T. Komninos, Compt. rend., 1918, 16’7, 781 ; Bull. SOC. chim., 1918, [iv],20, 449 ; A., i, 6.l4 P. Karrer, H e b . Chirn. Actu, 1919, 2, 89 ; A., i, 16094 ANNUAL REPORTS ON THE PROGRESS OF CHEMISTRY.A crystalline intermediate product containing chlorine but notbromine is collected and boiled with water.The reactions involvedappear to be:(i) CBrN + HC1 = CHBrtNCl; (ii) CHBr:NCl + C,H,(OH),=C6H3(OH),*CH:NC1 + HBr ; (iii) C6H,(OH),*CH:NC1 xrC,H,( OH), CH 0.Acetyl chloride and styrene in the presence of stannic chloridegive P-chloro-P-phenylethyl methyl ketone, which in its t'urn yieldsstyryl methyl ketone on treatment with diethylaniline.15In the course of an investigation 16 of the condensation productsof o-phthalaldehyde with dimethylaniline both normal and abnormalproducts were isolated. With excess of dimethylaniline and zincchloride the leuco-base of o-phthalaldehyde green,ctiH*[c.H (%H4 'NMez)z]z 2was obtained, whilst if the dimethylaniline was restricted to twomoleoular proportions o-aldehydoleucomalachite green was the pro-duct. When, however, o-phthalaldehyde and dimethylaniline werecondensed on the water-bath by means of concentrated hydrochloricacid a red base, C2,H,,0,N, was isolated, and convincing evidenceis available that this substance must be regarded as 2-o-aldehydo-phenyl-3-p-dimethylaminophenylindone :c a o *C,H,--G* 70NMe,*C6H4-C*C6H4The AT-alkyloximes have frequently been formulated as cyclicethers, thus, R-CH-NR, but just as the azoxy-compounds are now\/0regarded as containing the group *N:NO-, so the substances underconsideration may have the constitution CHR:NO*R.H.Staudinger and K.Miescher17 adopt this view, and also thename ' nitrone ' first suggested by Pfeiffer.l8 It is now found that'keto' nitrones, CR,:NO*R, are readily obtained by the action ofaliphatic diazo-compounds on nitroso-compounds, possibly in accord-ance with the scheme:NR*NO+CR,<j.J --+ NR<'--R -+ NR<? CR,:NR:O.The presence of two double linkings in the nitrones is renderedCR,*N CR,16 G. Langlois, Compt. rend., 1919, 168, 1052 ; A., i, 332.l6 E. Weitz, Annalen, 1919, 418, 1 ; A., i, 290.17 Helv. Chim. Acla, 1919, 2, 554 ; A., i, 584.lY Anncclen, 1916,411, 72 ; A., 1916, i, 327ORGANIC CHEMISTRY. 95tolerably certain by the fact that they combine with diphenyl-keten in two stages, thus:-0-CPb,:NPh:O + CPh,:CO + CPh2:NPh<Cph2>C0 (I)Diphenyl-N-phenylnitrone (from nitrosobenzene and diphenyl-diazomethane) is reduced by iron powder to benzophenoneanil, andoxidised by ozone to benzophenone and nitrobenzene. Boiling dilutesulphuric acid hyclrolyses it t o benzophenone and p-aminophenol.By heating the compound (I) a t 190° tetraphenyl-N-phenylnitrene,CPh,:NPh:CP&, is obtained, and constitutes the first example ofan entirely new type of substrance derived from the hypothetical' nitrene,' CH,:NH:CII,. This compound crystallises in pale yellowprisms melting a t 137'. On reduction with aluminium amalgam ityields dibenzhydrylaniline, NPh(CHPh,),, which was synthesised forcomparison.Many other nitrones and a few nitrenes have beenprepared and their properties examined in detail.Sub s t it u t ion and Orient at ion.It is not possible to notice the greater part of the systematicwork falling under this head, but it should be stated that therehas been considerable activity in this field, and many gaps havebeen usefully filled.On iodination 19 with the required quantity of iodine and nitricacid , iodobenzene gives pdi-iodobenzene ; chlorobenzene givespchloroiodobenzene, and bromobenzene, pbromoiodobenzene.pChloro- and pbrorno-toluenes give pchloro- and p-bromo-benzoicacids respectively, the methyl groups being oxidised to carboxyland no entry of iodine taking place.From benzoic acid, miodo-benzoic acid was obtained, and from o-phthalic acid, 4-iOdO-O-phthalic acid. Phenylacetic acid gives p-iodophenylacetic aoid andcinnamic acid, p-iodocinnamio acid.This method had previously been employed by G .M. Robinsonin the iodinat.ion of 0- and pnitzoanisole.m e interesting case of the nitration of benzotrichloride has beeninvestigated both by Vorlander 19@ and by E. Spreckels.20 Inthe very careful work of the latter precautions were taken to avoidI@ R. L. Datta and N. R. Chatterjee, J . Amer. Chem. SOC., 1919, 41, 292;A., i, 153.2o Ber., 1919, 52, [B], 315 ; A., i, 263.19a LOC. C i t 96 ANNUAL REPORTS ON THE PROGRESS OF CHEMISTRY.hydrolysis of the trichloride, and nitrobenzotrichlorides were ob-tained. Nitrogen pentoxide in carbon tetrachloride a t - loo yieldsits main product m-nitrobenzotrichloride with about 20 per cent. ofthe p-derivative and some ortho.Benzoic acid yields a far higherproportion of m-nitrobenzoic acid.The m-directive character of the ammonium salt group is alreadywell recognised, but D. Vorliinder and E. Siebert 21 have made adefinite contribution to the subject in demonstrating that onbrominating phenyltrimethylarnmonium bromide the m-bromo-deriv-ative is obtained, and also that m-nitrophenyltrimethylammoniumnitrate results from the nitration of phenyltrimethylammoniumnitrate. Neither reaction proceeds a t all readily, in harmony withthe usual experience of m-substitutions.The relations of the nitro-derivatives of diphenylarnine have beenelucidated,22 and the following scheme23 illustrates the course of thereaction between nitric acid, nitrous acid, and diphenylamine a t theordinary temperature, and a t low concentrations of the interactingcompounds :DiphenylamineJ.DiphenylnitrosoamineIJ. J.4-Nitrodiphenylnitrosoamine (2-Nitrodiphenylnitrosoamine)4-Nitrodiphenylamine (2-Nitrodiphenylamine)! I 1 I + J .J . J.4 : 10-Dinitrodiphenyl- 2 : 10-Dinitrodiphenyl- (2 : 8-Dinitrodiphenyl-nitrosoamine nitrosoamine ni trosoamine)4 : 10-Dinitrodiphenyl- 2 : 10-Dinitrodiphenyl- 2 : 8-Dinitrodiphenyl-amine amine amineI + I IJ.I + J.2 : 4 : 8-Trinitrodiphenylamine (2 : 4 : 10-Trinitrodiphenylamine)I I I I2 : 4 : 8 : 10-Tetranitrodiphenylamine. J. . +The compounds shoswn in brackets have not been isolated, but21 Ber., 1919, 52, [B], 283 ; A., i, 320.22 H. Ryan and T. Glover, Proc.Roy. Irish Acad., 1918, 34, [B], 97 ; A.,13 H. Ryan and P. Ryan, ibid., 1919, 34, 212 ; A., i, 482.are probably present in some of the fradions obtained.i, 13ORGANIC CHEMISTRY. 97LTatural Products.Guuiayefic A cid.-The communication 24 under review coiistitutesa notable advance in our knowledge of the constituents of resins.On dry distillation of guaiacum resin, two substances of unknownconstitution are produced, namely, guaiene and pyroguacin orhydroxymethoxyguaiene. Guaiene is now proved to be 2 : 3-dimethylnaphthalene, which was synt-hesised by a method thatleaves no doubt as to its constitution. P-Phenylisopropyl alcoholwas converted into the corresponding bromide and then into ,ethylP-phenylisopropylmalonate, CH,Ph*CHMe*CH(CO,Et),, by con-densation with sodiomalonic ester.This was methylated by theusual method, and the dibasic acid obtained by hydrolysis furnishedy-phenyl-aP-dimethylbutyric acid, CH,Ph*CHMe*CHMe*CO,H, onbeing heated a t 170-190°. Kipping's method was then requisi-tioned in order to close t*he naphthalene1 ring, the acid chloride ofthe above acid being treated with aluminium chloride so. as t oobtain l-keto-2 : 3-dimethyl-1 : 2 : 3 : 4-tetrahydronaphthalene (I).(11.; (111.)The corresponding alcohol, obtained on reduction, loses watel a t200°, yielding the dihydronaphthalene derivative (11), and thedibromidei of this is convkrted into 2 : 3-dimethylnaphthalena bythe action of alcoholic potassium hydroxide. Guaiene was a t firstthought to be 1 : 2-dimethylnaphthalene, and the latt4er substancewas also synthesised by somewhat similar methods.The inter-mediate was, in this case, the ketone (111), and the second methylgroup was introduced by the action of magnesium methyl iodide.Guaiaretic acid, the isolatioii of which from the resin ipdescribed, has the formula C,,,H%O,, is optically active andunsaturated. Its dimethyl ether, C,,H,,(OMe),, may be reducedunder vigorous conditions to a dihydro-derivative, isolatedboth in an actlivei and inactive form. The latter is thoughtio be a nzeso-modification, and the conclusioii is drawn thatthe molecule of the hydro-derivative contains two asymmetriccarbon atoms symmetrically disposed. Guaiaretic acid methylether yields verat<ric acid on oxidation with potassium per-24 C : .Schroeter, T,. Lichtenstutit, anti 13. Trineii, B e y . , 1918, 51, 1587 ; A . ,i , 84.REP.-VOT,. XVI. 98 ANNUAL REPORTS ON THE PROGRESS OF CHEMISTRY.manganate, and at the same time a portion of the substance isactually reduced t o hydroguaiaretic acid methyl ether. The ex-planation of this quite remarkable transformation is found in theaction of Hubl’s iodine solution on guaiaretic acid methyl ether.whereby i-hydroguaiaretic acid methyl ether and dehydroguaiareticacid methyl ether, C,,H,,O,, are obtained in the proportion of 1 : 2.It is probable, therefore, that, in the permanganate oxidatioii aportion of the guaiaretic acid methyl ether is reduced a t the expenseof another portion, and that hydroguaiaretic acid methyl ether isisolated on account of its great stability towards oxidising agents.i-Hydroguaiaretic acid methyl ether yields dibromo- and dinitro-derivatives, and the latter on reduction is changed into a diamine,which could not be resolved with the aid of d-tartaric acid.Guaiaretic acid methyl ether must, in view of thel above andother evidence, have the constitutionC,H3(OMe),*CH:CMe*CHMe*CH,*C,H,(OMe),,and guaiaretic acid is thus clearly related t o eugenol.Capsaickn.-The pungent principle of cayenne pepper is a snb-stance of general interest on account of its remarkable physiologicalproperties. A great step forward has been made in regard to theelucidation of its chemistry, and, indeed, prior t o the investigationsof E.K.Nelson 25 and of A. Lapworth and F. A. Royle,2G nothingwas known beyond t-he most elementary details. The hydrolysisof capsaiciii, C,,H,,O,N, by means of methyl-alcoholic hydrochloricacid yields 4-hydroxy-3-methoxybenzylamine (I), prepared for coin-M~O’)CH,-NH,H 01 \/parisoii by the reduction of vanillinoxime. The acid fragment isbest obtained by the use of 25 per cent. sodium hydroxide at 180O.It is found to be a new decenoic acid, C,,H,,O,, yielding by reduc-tion a decoic acid not identical with 12-decoic acid from coconut oil.Nelson therefore concludes that capsaicin is an amide of the consti-tution 11.Lapworth and Royle, who made a careful study of the isolationand properties of capsaiciii, obtained veratric acid by the oxidationof capsaicin methyl ether.Further, the vigorous reduction ofcapsaicin by means of sodium and alcohol was found to yieldammonia and a fatty alcohol boiling a t 216--217O, and convertiblez6 J . Amer. Chem. SOC., 1919, 41, 1115 ; A . , i, 543.26 T., 1919, 115, 1109ORGANIC CHEMISTRY. 99by oxidation into n-nonoic acid; also the action of inorganic acidchlorides on capsaicin gave a nitrile, which was changed byhydrogen peroxide and dilute sodium hydroxide at 40° into anainide melting a t 98-100°, which is the melting point of thesrnide of n-nonoic acid. There is, therefore, still some doubtas to the nature of the fatty chain and its mode of attachment tothe vanillylaminel moleculel. Lapworth and Royle originally sug-gested a dihydro-oxazole constitution, and, in a note attached totheir communication, express the1 opinion that this possibility is notwholly excluded as the result of Nelson's work on the hydrolysisof capsaicin.A somewhat allied topic is the pungency of synthetic compoundsrelated to zingerone, and this has been investigated and certaingeneralisations have been made .27 o-Hydroxystyryl methyl ketonewas found t o be exceptionally pungent.Tropic A cid and Truzillic ,4 cids.-Although not strictly naturalproducts, it is convenient t o mention at this stage that much atten-tion has been paid to these and related subjects during the pastyear.The preparation of tropic acid has been simplified,2g and themethod regarded as most economical is the following. Acetophen-one was converted into atrolactinic acid by the cyanohydrin method,and the latter, by distillation under diminished pressure, gaveatropic acid, which was transformed into P-chlorohydratropic acidby the' action of hydrogen chloride in ethereal solution : this, inturn, was hydrolysecl by aqueous sodium carbonate, and tropic acidobtained.Tropic acid has been resolved by 13.King 29 and also by McKenzieand wit'h almost identical results. The isoatropicacids (a and @) are obtained by heating atrolactinic acid in anatmosphere of carbon dioxide. Concentrated alkalis convert thea-acid into the @-acid, and it is therefore probable that theisomerism of these dimeric atropic acids is stereochemical.Adequate arguments have been put forward31 in favour of theview that the acids have the formula I, and are cis-trans-isomerides.2.7 (Mrs.) L.K. Pearson, Pharrn. J., 1919, 103, 78 ; A., i, 489.?* A. McKenzie and J. K. Wood, T., 1919, 115, 828.29 Ibid., 476.31 L. Smith, Lunds. Univ. Awskr., 1919, [ii], 14, 3 ; A., i, 486.E 230 L O C . cit100 ANNUAL REPORTS ON THE PROGRESS OF CHEMISTRY.The truxillic acids (a and P ) may also be stereoisomerides havingbhe formula 11, since H. S t ~ b b e ~ ~ has shown that a-truxillic acidP h CO,H\/ cCyHPh*QH* CO,HC H Ph CH C0,Hyields truxone by the action of sulphuric acid only by virtue of aninitial depolymerisation. The absorption curves of a- andP-truxillic acids exhibit close similarity. Stobbe considers thattruxone is C27H1S03, a conclusion reached by a consideration of itsrelations with truxene and tribenzoylenebenzene, C,7H,,0,, butR.St'oermer and G. Foerster33 prepared a methyl ether of thedioxime of a-truxone, and the result of a molecular weight deter-illination in benzene gave the formula C,,H,,O, for truxone.R. Weissgerber and 0. Kruber 34 have performed a remarkabletour de force in isolating four pure dimethylnaphthalenes from theheavy oil coal-tar fraction boiling at9 360-265O.1 : 6-Dimethylnaphthalene is that isomeride which is sulphonaiedmost readily in the cold. Its sulphonic acid was isolated andhydrolysed by steam a t 130-140°. The constitution of the liquidhydrocarbon was proved in several ways, for example, by oxidationto the dicarboxylic acid, which was synthesised in stages fromP-naphthylamine-5-sulphonic acid.2 : 6-B~methylnaphthtulene.-Sulphonation at 135-140O convertsthe 1 : 6-isomeride into soluble products and yields a sparinglysoluble 2 : 6-dimethylnaphthalenesulphonic acid.The hydrocarbonobtained on hydrolysis melts a t llO-lllo, and is identical with thedimethylnaphthalene obtained by Baeyer and Villiger 35 fromionone.2 : 7-Dimethylnaphthalene .-This new isomeride is isolated byremoving as much of the 1 : 6- and 2 : 6-isomerides as possible; a32 Ber., 1919, 52, [B], 1021 ; A., i, 329.33 Ibid., 1255 ; L4., i, 444.34 Ibid., 346 ; A., i, 315.35 Ibid., 1899, 32, 2429 ; A., 1 S99, i, 921ORGANIC CHEMISTRY. 101sulphonation a t 40° of recovered hydrocarbon then gives a pastymixture of acids, which is crystallised from 30 per cent.sulphuricacid. On hydrolysis, 2 : 7-dimethylnaphthalene is obtained (m. p.96--97O), and its constitution was determined by the, usual methods.2 : 3-Dimethylnaphthalene (guaiene, see above) was obtained 36 inrelatively small amount from the soluble sulphonic acids accom-panying the 2 : 6-dimethylnaphthalenesulphonic acid.The results of the work of It. Pummerer and E. Cherbuliez 37 onthe oxidation of l-methyl-@-naphthol are of much interest, but theoriginal must be consulted, as the investigation is too complex tobe1 usefully summarised.An interesting and unexpected observation 38 occurs in thePatent literature'. 1 : 6-Dihydrosynaphthalene is condensed withphthalic anhydride in the presence of boric acid to 1 : 6-dihydroxy-o-naphthoylbenzoic acid, which has a very sweet tastel.The corre-spopding 1 : 5-compound is tasteless.The action of bromine on juglone (I) in hot acetic acid leads tothe formation of a tribromojuglone (11), which is a brilliant redcompound, and constitutes, it is claimed,3Q a naphthalene dye of anew type.0/\AI l l\/\/HO 0Br 0The substance dyes cotAton iikordantecl with tannin in ecru shades,whilst its indigo-blue sodium salt dyes wool and silk directly.Very little of importance has been published during the yearunder review on the chemistry of anthracene, phenanthrene, andhigher polynuclear hydrocarbons.Alicyclic Group.I n 1915, Beesley, Ingold, and Thorpe40 showed that a cyclo-propane ring in the spiro-position wit,h respect t o a cyclohexanering was more readily formed than a simple cyclopropane derivative36 R.Weissgerber, Ber., 1919, 52, [B], 370 ; A., i, 318.37 Ber., 1919, 52, [B], 1392, 1414; A., i, 439, 442; R. Pummerer, Bey.,38 Society of Chemical Industry in Basle, D.R.-P. 311213 ; A., i, 403.39 A. S. Wheeler and J. W. Scott,, J . Amer. Chem. SOC., 1919, 41, 833 ; A.,4O T., 1015, 107, 1080.1919, 52, [B], 1403; A., i, 440.i, 490102 ANNUAL REPORTS ON THE PROGRESS OF CHEMISTRY.of closely analogous character; also the group I1 was, when formed,more stable than the group I.The conclusion drawn was that a part of the strain on thecyclopropane ring is taken up by the cyclohexane ring, or, in otherwords, that, owing t o the cyclohexane valency angle exceeding thenormal, the valencies a and h enclose a smaller angle than thecorresponding valencies c and d .A similar, but far more complex,study has now been publisheld41 of the va!ency stabilities in com-pounds containing the skeletonsc, “,/C-‘--C ,c -c\ Y,,CP- c>q I lb \c-C’ \c-c >C( 111 and c< c/ ‘C-GFrom the theoretical discussion, it was deduced that ths bolld xshould be distinctly more stable than y, aiid a slight increase ofstability was anticipated in the case of the bond a as comparedwith /3. On the other hand, bond a should be slightly less stablethan 6, although this effect, being of the third order, might proveincapable of detection. Experimental results justified all theaepredictions, but this highly interesting papetr cannot be adequatelycondensed, aiid the reader is referred t o the original for details.A new bicyclic terpene, CIOHIG, which yields pinene nitsoso-chloride with amyl nitrite and hydrochloric acid, has been dis-covered 42 in Finnish turpentine.A new sesquiterpene has beenisolated 43 from a distillate obtained during the manipulation ofpine resin.Active pineiie nitrosochloride has been prepared 44 from themother liquors from which the usual inact’ive compound hasseparated. By heating with aniline, d-pinene was regenerated.By applications 45 of the method of ozonisation, the formula I isconfirmed for d-fenchene (Wallach’s D-l-fenchene), whilst the ex-pression I1 may be assigned with certainty to /3-fenchene (Wallach’snd-f enchenei and Semmler’s isof enchene) .The fenchene, boilinga t 145--147O, is probably 111, whilst the fenchene of lowest boilingdl C. K. Ingold and J. F. Thorpe. T., 1919, 115, 320.42 0. Aschan, Technikern, 1918; A., i, 336.43 0. Aschan, Finska Kem. Medd., 1918 ; A.. i, 338.44 E. V. Lynn, J . Amer. Chem. Soc., 1919, 41, 361 ; A., i, 212.45 R. H. Rosohier, Acad. Sci. Pennicae, 1919, [A], 10, 1 ; A., i, 408ORGANIC CHEMISTRY. 103point (Semmler's isoallofenchene) is probably IV mixed withAschan's P-pinolene (V), the' lather in relatively small proportion.CH,-CH-CH, 3le2C--CH-OH, Me,C--CH-CHC H,-CH-C: CH, CH,-UH-C:CH, CH,-CH-UMe(1.) (11.) (111. )I I I I y e 2 I I YH2 I I YH2 IiThe work of Windaus on cholesterol has been continued, andalthough these investigations cannot yet be usefully summarised, itis considered46 that the constitution of cholesterol has beenelucidated to the extent indicated in the1 expression/\ IIf this is subsequently confirmed, cholesterol will be the firstnatural product shown t o belong to the spiro-ring type.Aromatic Selenium Compounds.Aniline selenate does not yield the selenium analogue ofsulphanilic acid on being heated, but sn-substituted aromaticselenium compounds can be obtained 47 from phenylselenious acid(I).Nitration by sulphuric acid and potassium nitrate yieldsPn-nitrophenylselenious acid (II), which, on reduction with sodiumhydrogen sulphite, becomes di-m-nitrophenyl diselenide (111).The corresponding diamine (IV) may also1 be obtained fromm-nitroaniline by way of its diazonium derivative and m-nitro-46 A.Windaus and 0. Dalmer, Ber., 1919, 52, [B]. 162 ; A , i, 203.F. L. Pyman, T., 1919,115, 166104 ANNUAL REPORTS ON THE PROGRESS OF CHEMISTRY.plieiiyl seleiiocyaiiste.with tin and hydrochloric acid.The latter yields the base (IV) 011 reductioiiSe0,H Se0,H Se-Se Se-Se/\ /\ /\ /\ /I I 1N02 NH,\I \/R. ROB,INSON.PART I11 .-HE: T ER o c Y c L I c D I v I s I ON.THE work of the current year in this field has been somewhat dis-appointing, and the dearth of interesting material accounts for thebrevity of the present Relport. No new lines have been opened upon the grand scale of recent researches on chlorophyll and theanthocyanins, but, instead, a good deal of quiet progress has beenmade' in t h s field of alkaloidal chemistry.The tendency, noted inprevious Reports, towards the study of natural rather thansynthetic products appears still t'o hold, which is a matter forcongratulation.On the purely '' artificial " side, an interesting example of thebenzidine rearrangelment in the glyoxaline serie8 may be mentioned,further study of which might help to clarify our ideas concerningthe mechanism of that peculiar process. The conversion of isatininto a quinoline derivative is also of interest from the point ofview of theory.Apart from these, the1 interest in synthetic organic compoundsseems to have centred in the coumarin and indole groups, whichhave given rise to a number of investigations.The chemistry of natural products is 'represented by a study ofthe anthocyanins, with special reference to colour variation inflowers, and a series of important facts have been brought t o lightin this section of the subject.Steady progress is being made inthe examination of the alkaloids, especially in clearing up theconstitutions of the more recently isolated members of the group.Tlw Rhodim Series.A study of the reactions of thiocyanoacetonel has revealed thefact that, this substance can give rise to several different hetero-cyclic compounds according t o tho reagents employed t o producecondensation, and it now seems established that previous investiga-1 J. Tcherniac, T., 1919, 115, 1071ORGANIC CHEMISTRY. 105tions in this field had led to erroneous conclusions.I n an attemptto prepare thiocyanoacetone, Hantzsch and Weber obtained asubstance which they supposed to be hydroxymethylthiazole, anda t a later date Hantzsch3 believed that he had produced amino-methylthiazole by the action of ammonia on thiocyanoacetone.Both these ideas are found to be mistaken.The reactions with which we are concerned a t present aresymbolised in the following scheme :8H-Yccl +- CH,*CO*CH,*NCS ?!!y 4C,H,ONS\/ 8-iso-MethylrhodimHCI cab*cN2 -ChIoro-4-met.hylthiazolea-MethylrhodimFor the compound now termed a-methylrhodim, Hantzsch sug-gested the structure$H--?CH,*C CO\./NHwhich makes it a derivative of thiazole. The properties of thesubstance, however, do not in any way agree with this formulation.For example, the compound shows no trace of ketonic properties,nor does it behave like an alcohol.Phosphorus pentachloride actson it without displacing oxygen, which appears t o negative theassumption that the oxygen atom exists in the ketonic form or inthe enolic stlructure derivable from the ketone. Hydrolysis of thecompound le8ads to decomposition products, which cannot be derivedfrom such a structure as Hantzsch proposed.It is now assumed that the condensation reaction takes place inthe following stages :CH,*C CiN -+ CH;C C:Nf:K2-? 5H-Y\OH\\02 A. Hantzsch and J. H. Weber, Bw., 1887,Annalen, 1888, %9, 7 ; A . , 1889, 413.p - 7\/+ CH,*C C:NH020, 3127 ; A . , 1888, 256.E106 ANNUAL REPORTS ON THE PROURESS OF CHEMISTRY.and it has been shown that this formulation of the end-product isin agreement with the actual propertie6 of the compound.With regard to the nature of isomethylrhodim and &methyl-rhodim, it is too early yet to state that their structures have beendefinitely established.From the fact thatl a-methylrhodim andP-methylrhodim are readily interconvertible, i t seems reasonable tosuppose that the &compound is a polymeride of the other, and thatthe structures of the two are similar. On the other hand,isomethylrhodim has a chemical behaviour different from either ofthe other two substances, and it appears to be expressed most satis-f actorily as a polymeride of the following :fiH--TCH,*C coThe Pyraxoline Group.When phenylhydrazine is allowed to act on phenyl styryl ketone,distyryl ketone, or ethyl y-keto-A"'- pentadiene-a€-dicarboxylate, thephenylhydrazones, which are the first products of the reaction,become spontaneously converted into pyrazolines.Further ex-amination shows that the reaction4 is a general one unless one ofthe following conditions is fulfilled, in which case the phenyl-hydrazone is stable and can be converted into the pyrazoline deriv-ative only by the employment of special processes : (1) The sub-stitution of pnitrophenylhydrazine for phenylhydrazine ; (2) thepresence of a halogen substituenb in the phenyl groups of bothketone and hydrazine; (3) the presence of a methoxy-group in theortho-position in the ketone.When the pyrazoline derivatives obtained by this reaction wereexposed to the influence of Rontgen rays, they exhibited intensefluorescence, not only in the solid state, but also in solution, theintensity of the fluorescence in the latter case being markedlyaffected by the nature of the solvent.The st'ructural conditionsnecessary for the production of this Rontgen ray fluorescence appearto be different from those demanded for the power of fluorescingunder light rays. For example, if the pyrazoline derivative con-tains a phenyl or carbonyl radicle in the positions 3 and ti, i tfluoresces with Rontgen rays, but fails to do so when these un-saturated groups are displaced by hydrogen atoms or aliphaticgroups. Under the action of daylight, however, this substitution4 F.Straw, Ber., 1918, 51, 1457 : A., i, 41ORGANIC CHEMISTRY. 107appears to be insufficient t o destroy the fluorescent power, as suchcompounds fluoresce quite clearly even in diffused daylight.The assumption here made that the pure pyrazoline derivativesare fluorescent may, in the end, prove to be erroneous, as somecases have now been investigated wherein pyrazoline compoundsevidently give rise to highly complicated and strongly fluorescentsubstances,5 and i t is possible that the phenomena described abovemay be attributable, not to the pyrazoline derivative, but ratherto its products.It has been shown that ketopyrazolines containing the structure-CO*CH <N H-, give strongly fluorescent solutions when dissolvedin alcohol containing a trace of hydrogen chloride, and the originof this fluorescence has been traced t o the formation of complexmaterials which resemble in their physical aspect the rhodaminedyes.Thus, when hydrogen chloride is passed into boiling methylalcohol in which ethyl 5-benzoyl-4-phenylpyrazoline-3-carboxylate issuspended, a crimson precipitate is produced which appears to havethe composition C,,H,,O,N,. From i t two other substances havebeen obtained which have the compositions C,BH~O,N,C1 andCBH,0,N4. The latter is a colourless compound, for which thefollowing structure has been proposed :yHPh*C :CPh*r*N:$!*CO,EtC0,Et.C: N---N*CPh: C --CHPhAnalogous results are obtained with some other pyrazolinederivatives.An interesting example of solvent effect has come to light 6 inthe pyrazoline series.When p-bromophenyl styryl ketone isheated with ethyl diazoacetate, an ester is formed which has thestructure (I), but i f the reaction mixture is diluted with lightpetroleum, the end-product is the ester (11) :7'h e Gly orca lin es .A curious abnormality has been detected in the reaction betweencliazonium salts and the glyoxaline derivatives.7 It appears to beestablished as a general rule that diazonium salts will react onlyE. P. Kohler and L. L. Steele, J . Amer. Chem. SOC., 1919, 41, 1105 ; A.,i , 557.E. P. Kohler and L. L. Steele, {bid., 1093 ; A., i, 530. ' R. G. Fsrgher and F. L. Pyman, T.. 1919,115, 217, 1015.E* 108 ANNUAL REPORTS ON THE PROGRESS OF CHEMISTRY.with those glyoxalines which contain a free imino-group and alsoa hydrogen atom, or a displaceable group (such as a carboxylradicle) in the 2-, 4-, or 5-position.Exceptions to this are, how-ever, found in the cases of 5-methylglyoxalin~4-carboxylic acid andglyoxaline-4-carboxylic acid ; for although the acids themselvesbehave normally and couple with diazonium compounds, the estersdo not react a t all. Up tot the present, no definite deductions canbe drawn from these results, butl it is suggested that the source ofthe abnormality must be sought in some mutual influence of theimino- and carbonyl radicles.I n the course of this investigation, a most interesting exampleof the benzidine rearrangement was observed. When 2-benzene-azoglyoxaline was reduced with stannous chloride.the main pro-duct was found to be 2-amino-4-p-aminophenylglyoxaline, a result.which can only be attributed to intramolecular rearrangement ofthe benzidine typa :NH-- NH NH"It must be admitted that the occurrence of the benzidine changei n the case of a fivemembered ring is extraordinary? but it ispointed out that the conjugation of the bonds in the glyoxalinering furnishes a certain parallel to that which is present in thestructure of hydrazobenzene.A Sy~it7~esi.c of p-Collidirie.In the coursel of some synthetic investigations in the quinineseries, a mode of forming P-collidine (4-methyl-3-ethylpyridine)has been discovered.5 As a first step, 2 : 6-dihydroxy-P-collidine isprepared, either by heating y-cyano-P-methyl-a-ethylglutaconimidewith hydrobromic acid or by condensing ethyl acetoacetate withethyl cyanoacetate in the presence of sodium and treating theglutaconic ester thus formed with sodium hydroxide.The next.stage in the process consists in converting the dihydroxycollidineinto 2 : 6-dichloro-~-collidine by the action of phosphoryl chloride.Finally, the chlorine atoms are removed by means of hydriodicacid ; monochloro-P-collidine is the first, product, from whichP-collidine is formed a t a further stage in the reaction.* L. Ruzicka and V. Fornasir, HeJu. Chim. Acta, 1919, 2, 338; A., i, 550ORGANIC CHEMISTRY. 109The Indole Group.When certain isatogens are heated under pressure with alcoholichydrogen chloride, they yield less intensely coloured isomerides.9It is suggested that the strongly coloured materials correapond withthe structural type (I) containing the1 pseudo-quinonoid grouping,whilst the new products have the linking (11) within the molecule:9 C coIt will be noted that if this view can be substantiated, the changecorresponds with the conversion of a five-membered ring into abicyclic structure containing an extra three-membered ring.The markgd difference in colour between indigotin and its diacetyldelrivative has apparently been accounted for by the proof that tholatter contains both the acetyl groups attached t o the nitrogenatam .loA iiumber of substituted indirubins have been prepared by meansof three different reactions,11 namely, (1) condensation of isatinswith indoxylic acid, (2) condensation of isatins with anilinoisatin,and (3) condensation of isatins in the presence of acetic acid withthe technical fusion of phenylglycine.The yield in the last caseseems good.A new and rapid method far extracting indican from indigo-yielding plants has been worked out .12The Coumarin Group.Among the cycloparaffins, it is well known that t3he stabilitiesof the five- and six-membered rings approximate closelyto one another, for in some cases the five-membered ring can beconverted by intramolecular rearrangement into the six-memberedt.ype and vice versa. A somewhat similar phenomenon has beennoted in the flavons and coumarin series, where the ring containsan oxygen atom in place of one of the methylene groups of theP.Ruggli, Ber., 1919, 52, [B], 1 ; A., i, 221.lo D. Vorliinder and J. v. Pfeiffer, ibid., 325 ; A., i, 225.l1 J. Martinet, Compt. rend., 1919, 169, 183 ; A., i, 457.l2 33. M. Amin, AgriC. Rea. Inst. Pusa, Indigo Publ., No. 5 ; A., i, 283110 ANNUAL REPORTS ON THE PROGRESS OF CHEMISTRY.cycloparaffin. Thus, when a benzylide!necoumarone of the gelneralstructure (I) is treated successively with bromine and potassiumhydroxide solution, it may be converted into the’ correspondingflavanone (11) :(1.1 (11.)Another branch of the same subject was opened up by the dis-covery that the removal of hydrogen bromide from substances ofthe general type (111) may take place in either of two ways, result-ing in t4he one case in t-he production of a flavone derivative (IV)and in the other in the synthesis of a coumarin compound (V):0 OH 0(Iv.1 (111.) (V. 1During the present year a study has been made13 of certainexamples of this type with the view of determining the effect ofsubstituents on the course of the reaction.From the results which have been accumulated, it appears as ifthe governing factor in the problem is the position of substituentsin the phenyl radicle which lies nearest the double bond in themolecule of the type (111). Thus, 2-acetoxyphenyl 4-methoxystyrylketone dibromide yields a coumaranonel derivative when treatedwith concentrated potassium hydroxide, whereas the isomeric2-acetoxyphenyl 2-methoxystyryl ketone gives a flavone compoundwhen similarly treated.This recalls to some extent the phenomena observed in the form-ation of coumarones from phenoxy-acetals714 in which case theinfluence of substituents is so great that i t may inhibit the reactionof coumarone-formation completely.Thus the compound (I) yieldsthe coumarone (11), but no such ring-formation takes place at allif a methoxy-group is inserted into the benzene ring in a positionortho t o the side-chain, as in (111).OMeA. - 0 /\--o p , - 0 I t I - + I ‘ , / \ A H \/)HZCH (OEt), CH\/‘H /CH,CH(OEt),(1.1 (11.) (111.)I3 J. Tambor and H. GubIer, HeZu. Ckim. Acta, 1919, 2, 101; A., i, 215.R. Stoermer, Anwlert, 1900, 312. 334; A., 1900, i, 650ORGANIC CHEMISTRY. 111A somewhat analogous investigation has been made with regardto the influence of substituents on the stabilities of variouscoumaranone derivatives. By treating the coumaranone derivativewith nitrophenylhydrazine, itq is found possible t o determinewhether or not a rupture of the heterocyclic portion of the mole-cule has taken place or not under these conditions.15 It appearsLhat the furan ring of the 1 : 1-dialkylcoumaranones shows greatstability, as in such a case hydrazone-f ormation occurs withoutrupture of the ring.Another example of a similar kind is t o be found in some recentwork on the o-allylphenols .I6 When o-allylphenol is submittedsuccessively to acetylation, bromination, and treatment withalcoholic potassium bromide, i t might be expected, from analogy toKostanecki’s syntheses, that the parent substance of the flavones,‘‘ chromene” (I), would be formed.Actually, however, the reactiontakes another course, and a five-membered coumarin ring is pro-duced (11). The reaction appears to be a general one.(1.1 o- Nlylphenol (11.)Since the coumaranones might be supposed to be capable ofenolisation, it is of some interest to find17 that both chemical andspectrochemical evidence tends to show that they are purely ketonicin nature, there being pracOically no enolic modification detectableeither by measurements of refractive indices or by titration withbromine.Three new methods for the synthesis of chroman and coumaran 18have been devised, zinc chloride being used as a condensing agentin each case. In the first method, phenol is condensed with achlorohydrin, and a poor yield of the required product is obtained.Better results are obtained by using phenyl y-hydroxypropyl ether(obtained by the action of trimethylene chlorohydrin on sodiumphenoxide) .When heated with zinc chloridel, this compound givesa 30.per cent. yield of chronian. By employing ethylene chloro-hydrin and sodium phenoxide, phenyl P-hydroxyethyl ether isformed, and gives a 25 per cent. yield of coumaran on heating with16 K. von Auwers and E. Adenberg, Ber., 1919, 52, [B], 92 ; A., i, 218.16 R. Adams and R. E. Rindfusz, J . Amer. Chem. SOC., 1919, 41, 648 ; .A,1’ K. von Aumers, Ber., 1919, 52, [B], 113; A . , i, 230.I* R. E. Rindfusz, J . Amer. Chem. &., 1q192 41, 665 ; A., i, 342.i, 340112 ANNUAL REPORTS ON THE PROGRESS OF CHEMISTRY.zinc chloride.The action of zinc chloride on phenyl y-bronio-propyl ether produces a 65 per cent. yield of chroman, whilst theanalogous reaction with phenyl P-bromoe'thyl ether leads to a3 0 4 0 per cent. yield of coumaran.The Conversion of Isatin into (8 Qzcindine Derivative.Another example of the change of a five-membered heterocycliccompound into one containing 8 six-membered ring is furnishedby the conversion of indoles into quinoline delrivatives. Thischange is brought about in some cases by the action of nitrowfumes, 2-cyano-2 : 3-dihydroindole-2-carboxylamide being thus trans-formed into 2-hydroxyquinoline-3-carboxylamide.~~ It has nowbeen found 20 that diazomethane possesses the power of effecting asimilar change.When this reagent is allowed to act on isatinsuspended in ether, 2 : 3-dihydroxyquinoline is produced.A New Heterocyclic T?ypc.It will be remembered that by means of Skraup's reaction it ispossible to fuse a newly formed pyridine ring on to an alreadyexisting benzene nucleus. I n the case of the formation of quinoline,an aromatic amine is treated with glycerol and sulphuric acid inthe presence of an oxidising agent such as nitrobenzene. Thisreaction has now21 been utilised in order to fuse a pyridine ringon to a coumarin nucleus, with the production of a nelw type oftricyclic compound in which all three rings differ in character, onebeing a pyrone ring, the central one a benzene ring, and the thirda pyridine nucleus.It has been found that! the1 reaction takes place with great readi-ness, so much so that i t is undesirable to utilisei aminocoumarinsa t all, the nitro-derivatives being sufficient ; and this naturallysimplifies the synthesig considerably.When 6-nitrocoumarin is treated with glycerol and ,sulphuricacid in the usual manner, condensation takes place, with the form-ation of t-he intermediate compound (I).This mipht then condenseintrainolecularly in either of two ways, as shown in the formulze:G. Heller and P. Wunderlich, Ber., 1914, 47, 1617 ; A., 1914, i, 863.*O G. Heller, ibid., 1919, 52, [B], 741 ; A.. i, 283.2' R. B. Dey and M. N. Goswami, T., 1919,115, 531ORGANIC CHEMISTRY. 1.1300(111.)Conclusive evidence is st.ill lacking as to which of these com-pounds is produced, but the balance of probability inclines towards(111).Such a substance would logically be termed $-1:8-&0-naphthoxazone.The chemical character of the $-naphthoxazones does not differmarkedly from that of other quinoline derivatives except in twopoints. In the first place, the $-naphthoxazones dissolve in hotalkali hydroxides, yielding substances of a deep colour, whichappear to be unstable acids formed by the opening up of the pyronering, since they regenerate the parent, naphthoxazone when treatedwith acids. Secondly, although the $-naphthoxazones are colour-less and form colourless salts with acids, yet their additive productswith alkyl iodides possess deep colours ranging from dark yellowto scarlet-red.When dissolved in water, these ammonium saltslose ttheir colour. From this it would appear that the ions derivedfrom the ammonium salts are colourless, whilst the non-ionisedmaterial is coloured . Further investigation of this phenomenonpromises interesting results, as the case is evidently the converse ofthat of phenolphthalein and other indicators, which are colourlessin the molecular condition but yield coloured ions in solution.The Flawone Series.This branch of the hete<rocyclic compounds has been less workedon recently, but some progress is to be noted. By investigations inprevious years, the constitution of scutellarein had been narroweddown t o two alternative possibilities, for it might be either5 : 7 : 8 : 4’- or 5 : 6 : 7 : 4/-tetrahydroxyflavone.It has now22 been22 G. Bargellini, Gazzetta, 1919, 4@, ii, 47 ; A., i, 645I14 ANNUAL REPORTS ON THE PROGRESS O F CHEMISTRY.shown, apparently, that the latter view is the correct one, so thatthe st,ructure of scutellarein is that shown below (I).The synthesis of datiscetin appears to have been acconiplished,23although, on account of the lack of material, i t has been impossibleto carry out the last step in the process, the demethylation of thetrimethyl ether of datiscetin. It seems clear, however, that thesynthetia 5 : 7 : 2’-trihydroxyflavanol is identical with the, trimethylether of datiscetin. The formula of the synthetic product is shownabove (11).A number of amino- and azo-derivatives of the flavone serieshave been prepared, and their properties have been examined. Theresults show that the amino-group exerts a stronger auxochromicinfluence than does the hydroxyl radicle.24The Anthocyanins.I n the earlier stages of the investigation of the’ anthocyanins, thereduction of quercetin was shown to produce cyanidin, and in thisway the genetic relationship bet-ween the flavone and anthocyaninseriea was established .25 I n these researches, the reducing agentemployed was magnesium and hydrochloric acid, acting in theprmence of mercury.A further examination of this field has ledto most interesting results.% Instead of hydrochloric acid, organicacids have been employed t o act on the magnesium or zinc whichis used in the reduction of the flavonel derivative, and in this modifi-cation of the method certain complex salts are produced whichappear to throw light on the problem of plant colorations.For example, when myricetin (I) is reduced by this method ityields green-tinted compounds which have the compositionApparently the reaction proceeds in stagw, the phenopyryliumderivative (11) being formed first, and then passing by eliminatioii2* G.Bargellini and E. Peratoner, Gazzetta, 1919, 49, ii, 64 : A., i, 547.24 M. T. Bog& and J. K. Marcus, J. Amer. Chem. Soc., 1919, 41, 83 ; A.,2s Ann. Report, 1914, 11, 138 ; 1915, 12, 156.26 K. Shibata, Y. Shibata and I. Kasiwagi, J . Amer. Chem. SOC., 1919, 41,C,BH,,O,*Mg*OAc,[Mg(oAc),l,.i, 169.208; A., i, 166ORGANIC CHEMISTRY. 115of acetic acid into (111), which finally unites with magiiesiuniacetate to produce (IV).When, instead of myricetin itself, aOHOAc Mg*OAc\/ OHrhamnoside derivative, myricitrin, is employed, the reactioii givesrise to a deep blue substance which contains four molecules ofmagnesium acetate.It will be remembered that hitherto the reduction of flavonederivatives has always given rise, to red materials. The apparentanomaly is explained by the fact that when these new green orblue reduction products are treated with hydrochloric acid, theyalso yield red compounds, the action of the hydrochloric acid bring-ing about the displacement of the group *Mg*OAc by a chlorineatom, with the consequent formation of the red oxonium chloride.It has been proved, however, that even when hydrochloric acid ispresent, in certain casw a green or blue material may be formedwhich contains the group *MgC1 instead of the radicle *Mg*OAc.The change of colour induced by the elimination of themagnesium atom from the substances is attributed by the authorst o two possible factors.I n the first place, the phenopyrylium rin116 ANNUAL REPORTS ON THE PROGRESS OF CHEMISTRY.contains one hydroxyl group more than does the correspondingoxonium salt, and in the second place, the magnesium atom isassumed to take part (by means of auxiliary va1encie.s) in complex-formation. I n support of this, it may be pointed out that thereduction of a glucoside (in which one of the hydroxyl groups ofthe compound is displaced by a sugar molecule) proves that thismasking of the hydroxyl tends t o shift the absorption band towardsthe violet, whilst with regard to the other factor, compounds con-taining the group -MgCl (which is supposed to be lem active thanMgOAc in complex-f ormation) have green instead of blue colours.The authors have thus been led to put forwardviews as t o the cause of colour variation in flowers.them, metallic complex salts of the following typeMX0Ithe followingAccording toare importanti _Jfactors in flower coloration and give rise to the “blue ” antho-cyanins.The metals which they contain are probably calciumand magnesium. The “ violet ” and “ red ” pigments are assumedt o be complex salts containing felwer hydroxyl groups than the“blue” ones have, or t o be mixtures of the “blue” compouiidswit<h some red oxonium salts which have been formed from the‘ I blue ” compounds by decomposition with acids.Experiments on the action between natural anthocyaiiins aiidsolutions of the salts of alkaline earth and heavy metals appear t ofurnish evidence in support of the authors’ contentions.,4 N e w Ii~ydrastin,ine Synth esis.37By the action of chloroJmethyl alcohol on homopiperoiiylamine(I) in ethereal solution, homopiperonylaminomethanol (11) isformed, and this, when treated with 10 per cent.aqueous hydro-chloric acid, yields dihydronorhydrastinine (111), from whichhydrastinine itself can be obtained. The following formulae showthe steps in the process:( 3 3 2 CH2(/\A O / \ ACH2<()I I yHZ CH2CI.0H CH2<01 I y H 2 \/ NH, ___ + \/ NH/CH2*OH(1.1 (11.)27 K. W.Rosenmund, Ber. Deut. pharm. Ges., 1919.29, 200 ; A., i, 280ORGANIC CHEMISTRY. 117(111.)The Cirkchona A ZkaZoids.Further progress has been made in this group, but the resultsare not yet published in full, so that it is impossible to give1 a com-plete account of the work which has been carried out.It has been found2* that, by means of palladous chloride in dilutesulphuric acid solution, it is possible t o reduce cinchoniiie,cinchonidine, and quinine t o the corresponding hydro-compounds.Some experiments have been made29 in coupling cinchona deriv-atives with diazobenzene and reducing the products.A claim is put forward30 that the cinchona alkaloids can now hebuilt up from quinoline and piperidine compounds.In this form,the claim is possibly correct, but as the material which the authorsemployed was obtained, not by synthesis, but from t'he degradationproducts of the alkaloids themselves, it is evident that the completesynthesis of the cinchona alkaloids is still unachieved .Some intramolecular changes in cinchonidine have beendes~ribed,~I and the decomposition products of P-hydroxycinchoninehave been investigated.32Hyoscine and Oscine.The complications introduced into the study of alkaloids by theexistence of spatial relations are well illustrated in the case of thehyoscines. Hyoscine occurs in two optically antipodal forms,cl-hyoscine and Z-hyoscine, and this year an investigation33 has beenmade with the object of determining the stereocheinical relationsof these compounds.Since the hyoscines are compounds built up from tropic acid andoscine, and since each of the latter occurs in two antipodal forms.it is evident that there are no fewer than eight isomerides possible:if racemic and partly racemic varieties are included.These may28 M. Heidelberger and W. A. Jacobs, J . Amer. Chem. SOC., 1919, 41, 817;29 G. Giemsa and J. Halberkann, Ber., 1919, 52, [B], 906 : A . , i, 34230 P. Rabe and K. Kindler, ibid., 1918, 51, 1360: A . , i, 33.31 E. LBger, Compt. rend., 1919, 169, 67 ; A., i, 451.s2 Ibid., 168, 404 ; A., i, 170.33 H. King, T., 1919, 115, 476, 974.A., i, 493118 ANNUAL REPORTS ON THE PROGRESS OB CHEMISTRY.be represented by the following symbols, in which T stands fortropic acid and 0 for oscine:Partial Opticallyracema t es .pure forms./ I * z-T-d-o5. dl-T-d-0 Ic/ 2. 1-T-1-0\//\3. d-T-d-06. dl-T-d-0-\ 4. d T-1.0Partialracema tes ./+8. d-T-dl-0\+/+Now, when I-hyoscine is hydrolysed with either acid or alkali, ityields I-tropic and dl-oscine. This excludes from the above listall the first six possible structures, since 1-6 contain either d-tropicacid or an active form of the oscine, so that for Z-hyoscine we areleft with 7 and for d-hyoscine we are driven to choose 8.When the problem of optically inactive hyoscine is considered, itwill be' found even more complex. As a starting point, there arefour possible varieties of the active forms :1.I-tropyl-I-oscine,3. I-tropyl-d-oscine,2. d-tropyl-d-oscine,4. d-tropyl-Z-oscine.Inspection of these will show that 1 and 2 would form an inactivecompound when mixed together, as would also 3 and 4. Further,these two inactive mixtures could not be identical with each other,owing to the different mode of linking between the right- and left-handed forms of the acid and base. Finally, a third inactivemixture might be obtained by mixing all four varieties together inmolecular proportions, and if combination took place between themthis would represent the production of a third possible type ofsubstance in the crystalline form.Now, i f the formation of an ordinary racemic compound beassumed, in which only two molecules combined to form the inactivecrystal, it might reasonably be expected that two different types ofcrystals would be obtained, corresponding with the two pairs (1 4- 2)and (3+4) above.On the other hand, i f all four varieties arecombined together in one crystal, then no second crystalline com-pound need be expected. I n actual practice, only a single form oft*he crystalline racemate is known, which may be taken as support-ing the combination of the four active forms into one crystal.King, on the basis of his investigations, suggests that the formulORGANIC CHEMISTRY 119of oscine is allied to that of tropine, and considers that it may havethe structureThe Alkaloids of the Pomegranate Tree.The difficulties which beset investigators in the field of alkaloidalchemistry are well illustrated by this group of compounds.I n1917, researches showed that a reform of the nomenclature wasrequired, whilst this year further facts have come to light whichapparently point to an isomerism depending on the spatial arrange-ment of groups about a tervalent nitrogen atom.It has been proved 34 that methylisopelletierine has the structure(I), and that i t can be obtained from conhydrine, which appearsto be (11). Closer examination of this reaction, however, showsthat, along with methylisopelletierine, a second base, db-methyl-conhydrinone, is produced.35 The two bases give different oximesand hydrazones, which excludes the idea that the difference betweenthem is due to keto-enolic desmotropy.Hess is therefore driven to suggest that the isomerism should beascribed to a different, sitluation of the methyl group in the com-pounds in question.I f the piperidine ring be supposed to beopened out and then placed in the plane of the paper, the followingformulze illustrate the conception :C0Et.C.H H *C.COEt COEt-OH H*d&OEt +I '+ + II I I+&*~c~e TtLe*N MwN+ +N*MeI[7H214a. 6.In the formula b it will be seen that t.he methyl radicle is supposedto be spatially adjacent t o the carbonyl group, whilst in formula nthe carbonyl and methyl radicles are on opposite1 sides of the ring.Steric hindrance is thus assumed t30 account for the fact that on0s4 K. Hess and A. Eichel, Ber., 1917, 50, 1192, 1386 ; A., 1918, i, 33, 34.35 K.Hess, ibid., 1919, 52, [B], 964 ; A., i , 345120 ANNUAL &E;YORTS ON THE PROGRESS OF CHEMISPRY.compound reacts more readily with semicarbazide than does theisomeric substance.A suggestion of this kind is not new,36 but the present caseappears t'o be differentiated from previous ones in that the isomerismis still preserved when methyl iodide is added on to each isomeride.Up to the prwent, i t has not been possible to bring about the con-version of methylisopelletierine into dl-methylconhydrinone. Themethylation of d-conhydrinone by means of methyl sulphate pro-duced a mixture of racemic methylisopelletierine and methyl-conhydrinone.COEtThe occurrence of isopelletierine, /-\NH, among the pome- \-/granate tree alkaloids has now been e~tablished.~~The Areca Xut A lknloids.The areca or betel nut contains numerous alkaloids, and it must,be confessed that the literature of this branch of chemistry containsan almost equally numerous flock of erroneous observations anddeductions.Nearly all the work which was summarised in lastyear's Report on this subject38 has now been shown to be erroneous.Taking the results in the order in which they were dealt with lastyear, the following must be noted. Arecaine was supposed to bedefinitely proved to be an X-methyl derivative of guvacine. Thisidea seems now to be abandoned. Guvacine appears to be 1 : 2 : 5 : 6-tetrahydropyridine-3-carboxylic acid, and not the 4-carboxylic acidas was supposed last year. This change entails a correspondingalteration in the formula of guvacoline, which is guvacine methylester. Arecaidine is the 1-methyl derivative of guvacine, and areco-line is arecaidine methyl ester.The supposed conversion of methyl-guvacine into the ethyl ester of guvacine by boiling with alcoholichydrogen ahloride turns out to be an error due to the employmentof impure materials.These errors have been frankly admitted by their authors, so it36 See Ladenburg, Ber., 1893, 26, 854 ; 1903, 36, 3694; 8., 1893, i, 442 ;1904, i, 92 ; but compare Wolffenstein, ibid., 1894, 27, 2615 ; A., 1894,i, 627. See also Groschuff, ibid., 1901, 34, 2974; A., 1901, i, 745; andScholtz, itid., 1910, 43, 2121 ; A., 1910, i, 634.37 K. Hess, Ber., 1919, 52, [B], 1005: A., i, 348.3* Ann.Report, 1918, 15, 107 ; K. Freudenberg, Ber., 1918, 51, 1668 ; A . ,i, 93 ; K. Hess and F. Leibbrandt, ibid., 1919, 52, FBI, 206 : A., i, 220 ; E.Winterstein and A. B. Weinhagen, Zidsch. physiol. Chern., 1918, 104, 45 ;A . , i, 171ORGANIC CHEMISTRY. 121may be assumed that 110 further coiitroversy will arise in this parti-cular region of-the subject.The Purine Group.By far the most iinportaiit work in the purine group duringrecent years is that done by Johnson and his collaborators, whichhas iiow reached the) eighty-eighth paper on the pyrimidines.39 Ithas been impossible to give even the most niodest account of thedetails of this vast investigation froin year to year, as the papersdo not leiid themselves t o condensation; but it seems advisable todirect attention here to the extraordinary fertility which this branchof the subject has exhibited in the hands of the investigat,ors, whohave made it their owii. The work done in this field and in thekindred one of the hydantoins represents one of the most fruitfulresearches in nioderii organic chemistry.Apart from this, the purine group has yielded but little ofinterest during the current year.Some work has been done 011hydurilic acid and s-diinethylhydurilic acid,40 but it calls for nop i tiadar comment.Our knowledge of cryptopine has been exhided considerablyduring the present year; but the paper 41 on the subject extends to110 less than seventy-eight pages, and even a summary of it would betoo long for inclusion in this Report.The reader is thereforereferred to the Transactions for further information.The bark of Croton gzcbouga has been investigated,&Z and from itan acid has been extracted which appears to be 4-hydroxyhygricacid, since on methylation it yields a mixture of betonicine andturicine. The properties of betonicine and turicine have beenstudied in detail:HO* H-vH2 HO*YE€-YH,CH, CH*CO,H CH, CH-CO\/ INMe,--O\/NMe4Hydrosyhygric acid. Betonicine and Turicine.39 T. B. Johnson and I. Matsuo, J. Awaw. C?Lem. SOC., 1919, 41, 782 ; T. B.4o H. Biltz and M. Heyn, B e y . , 1919, 52, [B], 1298 ; A., i, 491.41 W. H. Perkin, jun., T., 1919, 115, 713.4z J. A. Goodson and H. W. B. Clewer, ibid., 923.Johnson and L. A. Mikeska, ibid., 810 ; A., i, 498, 499122 ANNUAL REPORTS ON THE PROGRESS OF CHEMISTRY.Further investigations of the harmine group have led to the coii-clusion that the structures proposed for harmine in the past areincorrect.It is now suggested 43 that the following formulE repre-sent the conipounds better :Harmine.Harmaline.The details of the evidence on the matter do not lend themselves tosummarisation, and must be consulted in the original paper. Somesuggestions as to the possible mode of synthesis of the compounds inthe plant are put forward; the detaiIs of these are t o be found inthe Report on Physiological Chemistry in this volume.A study of the action of nitric acid on brucine** leads to thefollowing results. The steps in the process may be represented asfollows :C,,H,,O,N, + C,,H&,N, -+f&H190@3 + C,,H,iO,N3,HNO3.The last compound is cacotheline.I n its general behaviour itresembles the nitroquinones except for the fact that treatment withsulphurous acid changes it into a deep violet or green substance,whereas one would expect a less intensely coloured quinol to be pro-duced in this reaction. It seems now to be established that theviolet substance is not a reduction product, but is formed from thecacotheline by isomeric change. This invalidates the previous argu-ment against the nitroquinonoid structure of cacotheline, whichtherefore seems to be established.The action of diazonium compounds on various alkaloids has beenstudied, and it is found 45 that morphine is the only member of theopium alkaloids which yields a true dye in this way.The reductionof the dye failed to yield any aminomorphine. Methyl- and ethyl-morphine do not give dyes. Curiously enough, the physiological43 W. H. Perkin, jun., and R. Robinson, T., 1919, 115, 933.44 H. Leuchs, Ber., 1918, 51, 1375 ; A., i, 35.4 5 L. Lautenschlkiger, Arch. Pharm., 1919, 257, 13 ; A., i, 344ORGANIC CHEMISTRY. 123action of morphine is destroyed by the conversion into the colouriiigmatter.Oxidation of thebaine by means of hydrogen peroxide leads to theelimination of methyl alcohol and the production of a tertiary basewhioh has ketonic properties.46 It is supposed that this base isallied to codeinone, which contains one oxygen atom less; and it istheref ore termed osycodeinone.On reduction, it yields oxydihydro-codeinone, which is supposed to have the following structure:CHThe compound forms a hydrochloride freely soluble in water andsufficiently stable t o allow of the solution being sterilised. Thehydrochloride serves as a narcotic under the name of eukodal.The synthesis of cytisoline appears to have been accomplished,and it is thus proved to . be 2-hydroxy-6 : 8-dimethylq~inoline.~~Assuming this to be correct, Spath 48 suggests that the most probableformula for cytisine is :CH,*CHMeNH*CH/ \c-c ) c 4 .CH,--N ‘ 9 / \CH\CO----C H//The group of anhalonium or cactus alkaloids 49 has been examinedduring the current year.50 Anhaline has been shown to be identicalwith hordenine, so that its formula must beHO*C,H,-CH,*CH,*NMe,.Mezcaline has been synthesised, and proves to be p-3 : 4 : 5-trimeth-oxyphenylethylamine, C,H,(OMe),*CH,*CH,*NH,. These substances46 M. Freund and E. Speyer, Munch. med. Woch., 1917, 64, 380 ; A., i, 345.4 7 E. Spilth, Monatsh., 1919, 40, 93 ; A., i, 453.4 a Ibid., 15 ; A . , i, 4.51.4 9 See also this year’s Report on Physiological Chemistry.5 0 E. Spiith, Monatsh., 1919,40, 129; A.. i, 548124 ANNUAL REPORTS ON THE PROGRESS OF CHEMISTRY.do not, properly speaking, belong t o this section of the Report, asthey are open-chain compounds, but it seems well t o include themwith the rest of the alkaloids.Misc ellane ow.The nitration of 5diphenyldihydroacridine and the reduction ofthe nitro-derivatives to amines51 has led to the discovery of a newclass of dyes which have been termed carbazines. The amino-com-pounds act as leuco-compounds, and yield the dyes by oxidation.Some furt.her work has been carried out on the methyluric acids."?By t,he condensation of quinolinic acid with various polyhydricphenols, a series of dyes has been prepared which are analogous tothe phthaleins.537Hexacyanogen, CN*C <x:C(CN)>N, has been obtained by heat- N. C (CN) .~ing a mixture of cyanuric tricarboxylamide and phosphoric oxide ina vacuum to about 250°.54 When passed over a hot platinum wire itdecomposes into dicyanogen. Water decomposes it by the elimina-tion of the three cyanogen groups with the production of cyanuricacid. Hexacyanogen appears to be unattacked by chlorine or iodine,and seems to be indifferent towards hydrogen chloride.Iiivestigation of the action of potassium cyanate on Schiff baseshas led to the discovery that the reaction is markedly influenced byt,he nature of the base used.55 Thus benzylideneaniline reacts withpotassium cyanate, with the formation of a four-membered cycliccompound :CHPh:NPh + HCNO=CHPh<NH->CO, NPhwhilst benz~'1ideiieethylainine yields a six-mem bered ring owing totwo molecules of isocyanic acid taking part in the process:CHPhzNEt + 2HCNO= CHPh<NH-Co N tC:t*CO>NH*The four-membered ring is easily broken down, and yields on hydro-lysis a mixture of benzaldehyde and phenylcarbamide; but it hasbeen shown that this reaction cannot be reversed, since these twosubstances do not condense together to produce any yield of theuretidone from which they are formed.2, 315, 379; A., i, 551, 552.51 F. Kehrmann, H. Goldstein, and P. Tschudi, Helv. Clbim. Acta, 1919,5 2 H. Biltz and 31. Heyn, Ber., 1919, 52, [B], 768, 784; A., i, 292, 293.53 P. C. Ghosh, T., 1919, 115, 1102.54 E. Ott, Ber., 1919, 52, [B], 656 ; A., i, 260.5 5 W. J. Hale, J . Arner. Chern. Xoc., 1919, 41, 370 ; W. J. Hale and N. A.Lange, ibid., 379 ; A., i, 224ORGANIC CHEMISTRY. 125I n view of further developments in this field of research, thefollowing nomenclature for these f our-membered ring-systems hasbeen put forward:NHCH,/ \CH, CM A 3 H \*/ /y\\gH../' \NH/CH,Q ., z\CdUretidoiie. Uretidine. Uretc.N --CH< >CONHUre tine. Uretone.A new method of preparing pyrrole-black has been discovered .sf;Pyrrole is ,treated with the calculated quantity of a very diluteethereal solution of magnesium ethyl iodide, and air is 'then drawnt,hrough the liquid for twenty-four hours. The pyrrole-black isdeposited from the solution and exhibits a much inore intense tintt,han is show11 by samples prepared by other niet,hods.A bicyclic seleriiuin heterocyclic coiiipound,5ihas been obtained by treating two niolecules of yutinaphthylenedi-amine in pyridiiie solutioii with one molecule of selenious acid dis-solved in aqueous pyridiiie. Some reactions of the substance aredescribed.A new series of compounds, described as parazenes, has been pre-pared.58 Members of the series contlain two benzene (or similar)nuclei linked together by two nitrogen atoms, each of which isattached to two para-carbon atoms in the rings. Three possiblephases for such a structure are shown below. I n the case whereone of the phenylene groups contains a substituent, all three phaseswould be different ; but in symmetricallybhe parent coiiipouiid the phases (I) andA. Angeli an2 A. Pieroni, Atti R. Accad.5 7 0. Hinsberg, Bey., 1919, 52, [B], 21 ; A.,6 8 A. Angel. Brit. Pat. 121347 : A.. i. 98.A . , i, 134.substituted menibers or in(111) would be identical:/"\/ \/\N//\ /\- 1 1 (1(TIT.)fiincei, 1918, [v], 27, ii, 300;i, 226126 ANNUAL REPORTS ON THE PROGRESS OF CHEMISTRY.The parazenes are obtained by acting with a condensing agent onbenzene or naphthalene derivatives which contain an amino-groupand a halogen atom in the para-positioii with respect to each otherand subsequently reducing the hydrosy-parazene thus produced.The members of the series are coloured compounds which yieldcolouring matters suitable for dyeing.I n concluding this series of Reports on the Heterocyclic Divisionof Organic Chemistry which have now been written by him for anumber of years, the author is again coiiscious of the limitationswhich are imposed on a reporter by the nature of the subject, andalso by the considerations of space. He is only too well aware thatmany most interesting subjects have not been dealt with evencursorily in these Reports year by year. This has not been due t oany lack of appreciation on his part. Some important papers havedefied the process of summarisation altogether ; others have beenomitted because of limitations in the space which it is possible t oallot t o the Report. In no case has any paper been left unmen-tioned without due reason; but it has obviously been impossible toconvert the Reports into a mere catalogue of subjects which havebeen investigated during the year. With this apology, the authorbrings his work t o a conclusion.A. IV. STEWART