ORGANIC CHEMISTRY.PART ALIPHATIC DIVISION.Alcohols, Aldehydes, and Ketones.THE action of fused alkali hydroxides at 250-300" upon ethylalcohol is slow and several reactions appear to take place simultan-eously, sodium carbonate, oxalate and acetate, hydrogen, methane,and ethylene being produced. Acetaldehyde under similar con-ditions at 250" reacts to the extent of 90% in accordance with theequation C,H,O + NaOH+NaOAc + H2. Higher temperaturespromote the formation of methane and sodium carbonate :C,H,O + BNaOH-+H, + CH4 + Na,CO,. Acetone was foundto behave like acetaldehyde.1Several communications have appeared dealing with the structureof the acetylenic y-glycols. In one of these2 a study is made ofthe physical properties and reactions of pc-dibromo-pc-dimethyl-Ar-hexinene and of the product resulting from the action of phos-p horus t ri bromide on p E - dimet h yl- A7 - hexinene- PE - diol . The crys t a1 -line dibromide, m.p. 39", previously obtained by G. D~poni;,~ iscomparatively inert and appears to possess the ethylenic structureCMe,:CBr*CBr:CMe, and not Dupont's acetylenic structure. Theglycol on treatment with phosphorus tribromide gives a mixture ofproducts, amongst which is a less stable isomeric dibromide, m. p.46-48", which probably possesses the acetylenic formulaCMe,Br*CiC*CMe,Brand passes easily into the ethylenic isomeride. Other varieties arepresent, however, and the reaction is apparently complex. Theseconclusions are in essential agreement with the results obtainedin another investigation of the same problem,4 with the exceptionthat in the latter case the acetylenic dibromide, m.p. 4648", wasfound to be comparatively inert. The action of hydriodic acidon the glycol effects conversion into a crystalline di-iodide, C,H,,12,H. S. Fry and E. L. Schulze, J . Amer. Chem. Soc., 1926, 48, 958; A.,W. N. Krestinski, J . Russ. Phys. Chem. Soc., 1926, 58, 1067; A,, 1927,Compt. rend., 1911, 152, 197; A., 1911, i, 173.J. S. Salkind and M. P. Sigova, J . Russ. Phys. Chem. SOC., 1926, 58.1926, 710.442. See also Ber., 1926, 59, [B], 1930; A., 1926, 1121.1039; A., 1927, 44262 ANNUAL REPORTS ON THE PROGRESS OF CHEMISTRY.m. p. 75", analogous in structure t o the ethylenic dibromide,5 andthe investigation has been extended to include the action of hydro-bromic acid on one of the stereoisomerides of as-diphenyl-AB-butinene-as-diol.This gives a complex mixture of products,amongst which are two solid dibromides, a solid tribromide, aliquid dibromide and a liquid monobromide. A detailed study ofthe properties of each of these shows that the monobromide, forwhich the formula CH-cHPh>O I I is suggested, yields no di-bromide on further bromination, and the dibromides must thereforebe formed by means of an independent parallel reaction. To thedibromide, m. p. 114-1 15", the structure CHPh:CBr*CH:CPhBris ascribed, and the isomeric dibromide, m. p. 92-95", may be eitherCHPhBr*CiC*CHPhBr or CPhBr:CH*CH:CPhBr.Oleyl alcohol (n-trans-At-octadecen-a-01) has been converted viathe liquid bromine additive compound, which is then treated withsilver acetate in acetic acid, into the diacetate of oleicerin (n-trans-octadecane-atK-trio1). The free trio1 obtained on hydrolysis is asolid, m.p. 126.5". The corresponding n-cis-octadecane-awtriO1(elaidicerin, m. p. 92") may be prepared in a similar way fromelaidyl alcohol.'A further addition to the number of organic compounds con-taining fluorine has been made by the preparation of trifluoro-tert.-butyl alcohol from isoamyl trifluoroacetate by reaction withmagnesium methyl iodide, the fluorine remaining unattacked.The alcohol (m. p. 20.8") may be converted by treatment withphosphorus pentabromide, but not with concentrated sulphuricacid, into yyy-trifluoroisobutylene, which combines without loss offluorine with hydrogen bromide and with bromine.8A study of the absorption curves in ultra-violet light.of acetalde-hyde and paracetaldehyde solutions indicates the presence of anenolic form of the aldehyde, the proportion being 1 in 335 foracetaldehyde in sodium hydroxide, 1 in 530 in hydrochloric acid,and 1 in 1045 for paracetaldehyde in hydrochloric acid.9The enolisation of acetone has been investigated by means ofthe compound (CH,:CMe*O*),HgY2Hg0 obtained from freshly pre-cipitated mercuric oxide and acetone in the presence of potassiumhydroxide. It appears that the cnolisation of acetone is analogousCBrCHPh5 J. S. Salkind, B. Rubin, and A. Kruglov, ibid., p. 1044; A., 442.6 J. S. Salkind and A.Kruglov, ibid., p. 1052; A., 443; Ber., 1926, 59,7 E. Andre and (Mlle.) T. Frangois, Compt. rend., 1927, 185, 387; A., 957.8 F. Swarts, Bull. SOC. chim. Belg., 1027, 36, 191; A, 443.9 8. A. Schob, Cornpt. rend., 1927,184, 1452; A., 751.1936; A., 1926, 1121ORGANIC CHEMISTRY.-PART I. 63to that of ethyl acetoacetate, a maximum amount of the enolicmodification being formed in the presence of 0.037df-alkali, furtherincrease in the alkali concentration being without effect .loThe action of bromine (1 mol.) on paracetaldehyde a t low temper-atures gives bromoparacetaldehyde, m. p. 27.5", together with thedibromo-compound. The former decomposes when heated a t 130"into bromoacetaldehyde. The reaction between bromine (3 mols.)and paracetaldehyde is complicated by the action of hydrogenbromide on the aldehyde, which leads to the formation of tetra-bromobutaldehyde. In all these changes very slight evolutionof hydrogen bromide occurs and the mechanism of the reactionis said to depend on an equilibrium between paracetaldehyde andacetaldehyde, the latter of which becomes enolised under theinfluence of bromine to give vinyl alcohol.The very unstableah-dibromoethyl alcohol is then formed, which passes into tribromo-paracetaldehyde and hydrogen bromide. The liberated acid maycombine with acetaldehyde or vinyl alcohol, yielding a-bromoethylalcohol, which with a p-dibromoethyl alcohol gives a mixture ofmono- and di-bromoparacetaldehyde 11 in accordance with thetwo equations :2CH2Br*CHBr*OH + CH,*CHBr*OH + C6Hlo03Br2 + 3HBrCH,Br*CHBr*OH + 2CH,*CHBr=OH --+ C6Hl10,Br + 3HBrBromination of polymerised aldehydes at low temperatures hasbeen employed as a means of obtaining a-hydroxy-aldehydes.Thus parapropaldehyde with bromine (1 mol.) at -lo", followedby treatment with alcohol, gives the acetal of a-bromopropaldehyde,which when hydrolysed with water yields a-hydroxypropaldehyde.Similarly from the acetals of aa-dibromopropaldehyde and a-bromo-heptaldehyde, pyruvaldehyde and a dimeric form of a-hydroxy-heptaldehyde were obtained.12 A study of the complex reactionbetween magnesium ethyl bromide and a-bromo-aldehydes revealsa certain similarity in behaviour between the latter compoundsand acyl halides.For instance, a-bromoheptaldehyde and mag-nesium methyl bromide give a very small amount of y-bromo-P-octanol, together with methyl hexyl ketone, a tertiary alcohol(probably (3-methyloctan-p-ol), and an olefine, C9H18, which resultsfrom dehydration of the tertiary alcohol.Lead hydroxide anda-bromoheptaldehyde give heptoic acid in place of the expectedlo W. L. Evans and W. D. Nicoll, J. Amer. Chem. SOC., 1925, 47, 2789;A., 1926, 51.l1 A. Stepanov, N. Preobraschenski, and M. Schtschukina, Ber., 1926, 59,2533; A., 1927, 42. Compare also R. Dworzttk, Monatsh., 1925, 46, 253;A., 1926, 385.la R. Dworzak and P. PfiEerling, Monatsh., 1927,48, 251 ; A., 105564 ANNUAL REPORTS ON THE PROGRESS OF CHEMISTRY.aldehyde-alcohol, and it is thus probable that the methyl hexylketone is formed by molecular transformation during the actionwith the magnesium methyl bromide, the reaction thereafterresembling that with an acyl halide.13The chemical properties of glyceraldehyde and of dihydroxy-acetone are of special interest in view of the importance of thesecompounds both in physiological processes and as simple hydroxy-compounds related to the aldoaes and ketoses.A. Wohl's method l4of preparing glyceraldehyde has been improved and a detailedexamination of the dl-compound has been commenced.15 Duringthe isolation of dl-glyceraldehyde from its diethylacetal a viscousopaque syrup can be obtained which possesses unexpected properties.It is definitely enolic in character, and is unimolecular in aqueoussolution, whereas the crystalline variety is bimolecular and inaqueous solution contains only a trace of the enolic modification.The enol appears to be present also in an alkaline solution of crystal-line glyceraldehyde and, since this enol is identical with the enolicmodification of dihydroxyacetone, an explanation readily followsof the mechanism of acrose formation from glyceraldehyde, theessential reaction being the aldol condensation of glyceraldehydeand dihydroxyacetone.The transformation of bimolecular glycer-aldehyde into dihydroxyacetone in yields up to 49% has beeneffected by treatment with boiling pyridineI6 and on the basisof this and other evidence, such as the indifference of the acetatetowards phenylhydrazine and the non-formation of an isopropyl-idene ether, the formula ( O<cH.cH2.0H)2 is suggested.Thediethylacetal of glyceraldehyde, on the other hand, reacts readilywith acetone to form the isopropylidene ether.The action of hydrobromic acid in acetic acid on diacetylglycer-aldehyde gives the bimolecular acetobromoglyceraldehyde, m. p.QH*OH168-169", to which the formula ( O<CH.CH,.0Ac)2 QHBr is ascribed.Methyl alcohol and silver carbonate give the correspondingacetylated methylcycloacetal, from which bimolecular glyceralde-hyde methylcycloacetal, m. p. 158.5-159.5", may be obtained by1 YH-OMethe action of ammonia in methyl alcohol, (O<cH.cQoH),.The very close resemblance between these reactions and the corre-sponding properties of acetobromoglucose is further shown by thel3 A.Kirrmann, Compt. rend., 1927,184, 1463 ; A., 751.14 Ber., 1898, 31, 1796, 2394; A., 1898, i, 655; 1899, i, 11.l5 H. G. Reeves, J . , 1927, 2478; A., 1172.18 H. 0. L. Fischer, C. Taube, and E. Baer, Ber., 1927, 60, 479; A., 340ORGANIC CHEMISTRY .-PART I. 65transformation of acetobromoglyceraldehyde to bimolecular acetyl-glyceraldehyde, m. p. 118.5'. Some similar bimolecular compoundshave been obtained from glycollaldehyde (e.g., glycollaldehydemethylcycloacetal, m. p. 72"), and from dihydroxyacetone, whichgives the met hylc ycloacetal CH,. OH] , m . p . 1 3 1- 1 32 O . l7The elucidation of the particular type of polymerisation foundin the dimeric compounds may be expected to yield results ofimportance both for this and for other fields of investigation.A convenient method for the preparation of isopropylideiieethers depends on the use of anhydrous zinc chloride as condensingagent in dry acetone solution.18 The ethers thus prepared aremore stable than those obtained by the aid of acids, and of themany substances treated by this method only d-glucose showed anabnormal behaviour.The process is specially valuable as a meansof isolating dihydroxyacetone.A general method for the preparation of ketone-alcohols of thetype CHPh(OH)*COR is that of acting on phenylglycollamide or thecorresponding nitrile with magnesium alkyl halides. Secondaryalcohols of the type CHPhROOH are, however, always formed asby-products in the case of the nitrile.19 The preparation of theketone-alcohols CHPh(0H)-COR, in which R = ethyl, propyl,isopropyl, n-butyl, isobutyl, and benzyl, is described.It has beenfound that a-keto-alcohols, when heated in alcoholic solution at120-130" with a small quantity of sulphuric acid, undergo molecularrearrangement. The carbonyl group moves so as to take up aposition as near as possible to the end of the chain, the tendencybeing for the formation of an acetyl group. Thus butyrylethyl-carbinol is transformed into propionylpropylcarbinol. The sugges-tion is made that a similar transformation may play a part in thecourse of alcoholic fermentation.20A detailed criticism has been contributed of the formulaCH,(OH)*O*SO,Na assigned by E. Knoevenagel21 to the sodiumhydrogen sulphite addition compound of formaldehyde. Thestability of the addition compound towards oxidising agents, itsready reducibility, and its reaction with phenols to give sulphonicacids are all incompatible with the above formula and evidenceof the direct attachment of sulphur to carbon is provided.Ammonial7 H. 0. L. Fischer and C. Taube, Ber., 1927, 60, 1704; A., 857.l 8 Idem, ibid., p. 485; A., 338.lo M. Tiffeneau and (Mlle.) J. Levy, BUZZ. SOC. chim., 1925, 37, 1247; A*,1926, 71.A. Favorski, ibid., 1926, 39, 216; A,, 1926, 600.21 Ber., 1904, 37, 4060; A . , 1904, i, 1027.REP.-VOL. XXTV. 66 ANNUAL REPORTS ON TRE PROGRESS OF CHEMISTRY.reacts with formaldehyde hydrogen sulphite to give a compoundwhose properties show it to be aminomethanesulphonic acid andfurthermore formaldehyde bisulphite and ethyl acetoacetate may bemade to yield ethyl a-sulphomethylacetoacetate,COMe-CH( CH,*SO,H)*CO,Et,which is hydrolysed in alkaline solution to P-sulphopropionic acidand in acid solution to y-ketobutanesulphonic acid.It is shownthat the change is not explicable on the grpunds of initial fissionof the aldehyde bisulphite with the addition of sulphurous acid a ta later stage. The attachment of both hydroxy- (or amino-) andsulphonic groups to the same carbon atom confers a special labilityon each, which is responsible, in addition to other reactions, forthe ready loss of the latter group as sulphur dioxide. In the secondof the two papers referred to, a wealth of experimental evidence iscited to support the formulation of formaldehyde bisulphite a.ssodium hydroxyme t hanesulphonate, CH, (OH) *S 03Na .*2Carbohydrates.Monosaccharides, Lactones, and G1ucosides.-The revision of thestructural formulze of the most commonly occurring forms ofglucose, reported last year, has received wide acceptance, and thegeneralisation which accompanied this revision has passed intocommon use in the formulations applied to the carbohydrate group.Corifirmatory evidence of the nature of the oxide rings in derivativesof the normal and also of the labile or y-sugars has been furnishedfrom many sides, and this evidence has now assumed so convincinga character as to place the issue beyond any reasonable doubt.There is still need for reform in the nomenclature adopted to definethe structural relationships of the sugars, and a paragraph on thisproblem at the end of the present sub-section reports on some recentsuggestions.A comparative study of ten methylated lactones derived fromsimple sugars has shown that the rate of hydrolysis of five lactonesobtained from normal sugars is many times greater than in the caseof the remaining five lactones, which are related to the y-sugars.%The curves showing the change lactone acid reveal well-markeddifferences in the stability of the lactones as between the two types.Those derived from normal sugars are seen to be %lactones havingsix-membered rings, whereas those related to y-sugars are y-Iactoneshaving five-membered rings.Six of the series of ten lactones are22 F.Raschig, Ber., 1926, 59, 859; A., 1926, 699; F. Raschig and W.Prahl, Annalen, 1926, 448, 265; A,, 1926, 939.23 H. D. K. Drew, E. H. Goodyear, and W. N. Haworth, J., 1927, 1237;A,, 750ORGANIC CHEMISTRY .-PART I. 67crystalline and others give characteristic phenylhydrazides whichare crystalline. It is possible to diagnose by these methods whichmember of a pair of related lactones belongs to the &type and whichto the y-type, and since these substances are obtained from sugarsby a simple oxidation which does not involve degradation or otherprofound change, the ring-structure of a sugar derivative can bereadily determined.In many cases the lactones have been submitted to oxidativedegradation to the dibasic acids. The characterisation of thesedegradation products by direct comparison with authentic referenceproducts hss furnished confirmation of the structural formul=which had previously been a.llocated to the lactones on the basisof the physical studies outlined above.Examples of this kindmay be quoted. The crystalline sugar, tetramethyl glucose (I),obtained by methylation methods from a- or p-methylglucoside,yields a tetramethyl S-gluconolactone (11) which suffers degradationwith hot nitric acid. The chief product of this change has beenrecognised 2* as xylo-trimethoxyglutaric acid (111), which gives acrystalline methylamide identical with that prepared 25 by oxidisingthe %lactone (IV) obtained from the trimethyl xylose (V) which isderived from the usual form of methylxyloside.I $‘H*OH{H-OH 1 Q0,H .p , H H*T*OMe[FHoOMe]4 ? [QH*OMe], --+ [yH*OMel4 MeO-7-HCH, -A CH,OH C02H H*y*OMe 1(VI.1 (VII.) (VIII.) CH,-’(V. )It is conceivable that (111) might arise from the direct oxidationof a tetramethyl sugar of formula (VI), but this is much less likelyto be so if the intermediate stage is recognisable as a lactone, sincethe existence of an c-lactone is extremely doubtful. This possibility24 W. N. Haworth, E. L. Hirst, and E. J. Miller, J., 1927, 2436; A., 1173.2 5 W. N. Haworth and D. I. Jones, ibid., p. 2340; A., 105968 ANNUAL REPORTS ON THE PROGRESS OF CHEMISTRY.is, however, entirely negatived by the experimental observation 26that the monobasic acid of a substance corresponding to (VI) hasbeen prepared, namely, 2 : 3 : 4 : 5-tetramethyl gluconic acid (VII),and this does not form a lactone, nor does it oxidise to the dibasicacid (111); under the same conditions as those adopted for thedegradation of (11) to give (111), it passes into tetramethyl saccharicacid (VIII).It is represented that there can be no other interpret-ation of these results than that the tetramethyl glucose is correctlyformulated by (I) and that a- and p-methylglucosides possess a six-membered (amylene-oxide) ring. That the free a- and P-glucoseshave the same oxide-ring structure seems more than probable.Referring now to the two aldohexoses galactose and mannose, aconfirmation is given that the usual forms of methylgalactoside andalso a-methylmannoside correspond to the same structural classific-ation as the two normal forms of methylglucosides (a- and p-).Crystalline tetramethyl galactose (IX) yields on oxidation tetra-methyl 8-galactonolactone (X), the constitution of which is nowverified 27 by degradation to the Z-arabo-trimethoxyglutaric acid(XI), which is identical with that obtained 2* by oxidising Z-trimethyl6-arabonolactone (XII) and this in turn is obtained from the normaltrimethyl arabinose (XIII).IVH-OH -1Similarly also with cc-methylmannoside : this yields a tetra-methyl mannose (XIVj which passes to a crystalline tetramethyl8-mannonolactone (XV).Degradation of the latter leads to the2 6 W. N. Haworth, J. V. Loach, and C. W. Long, J., 1927, 3146.27 W. N. Haworth, E. L. Hirst, and D.I. Jones, ibid., p. 2428; A., 1173.28 W. N. Haworth and D. I. Jones, loc. cit.; 5. Fryde and R. W, Hum-phreye, J., 1927, 559; A,, 449ORGANIC CHEMISTRY .-PART I. 69isolation of d-arabo-trimethoxyglutaric acid (XVI,,29 which is theenantiomorph of (XI).p 2 HMeO*y*H 1 Me 0.F TO-1 -€I MeO-7-H--+ MeO-F*H Me0q.HMeg:v::Me I H-Q-OMe 1 H*Q=QMe? ! IH*‘i €I*F--J CO,HCH2-OMe CH,. ORfe(XIV.) (XV-) (XVI.)The authors conclude that t’he usual forms of methylgalacto-side and methylmannoside are normal forms in that they arestructurally similar to the normal methylglucosidea. Galactoseand one of the known varieties of mannose seem to possess, there-fore, the six-membered ring (amylene-oxide) structure. This canalso be said for the a- and p-methylarabinosides and methyl-xylosides , and similarly for ordinary arabinose and xylose.A like interest attaches to the determination of the structureof p-methylfructoaide through its characteristic derivative, crystal-line tetramethyl fructose (XVII).The conclusions reported lastyear30 have received extended verification in that this sugar isshown definitely to pass on oxidation with nitric acid to the amylene-oxide lactol-acid (XVIII) 31 and then, by degradation, t o d-arabo-trimethoxyglutaric acid (XIX), which is identical with (XVI).FH,*OMe , ---- QO2H d;-o-H(XVIII. CH2 )r--?*OH QO2H 1 Me0.y.H \ MeO-F*H Me 0 -y*HH*y*OMe ‘ H*r*OMe H-F-OMe L H*$?*OMe H*y*OMeCO,H(XIX.) Q g ; 2 M e (XVII. )It is thus seen that p-methylfructoside possesses the same typeof cyclic structure as a- and p-methylglucosides; and evidently thecrystalline fructose, in common with a- and p-glucose, is suitablyrepresented by the amylene-oxide formula.Turning now to the sugar derivatives related to y-methyl-glucoside (XX), it is found that the derived tetramethyl y-glucose(XXI) yields a crystalline tetramethyl y-gluconolactone (XXII) ,which passes by oxidative degradation 32 to d-dimethoxysuccinic20 E.H. Goodyear and W. N. Haworth, J., 1927, 3136.30 Ann. Reports, 1926, 23, 80.31 W. N. Haworth, E. L. Hirst, and A. Learner, J., 1927, 1040; A., 649.aa W. N. Haworth, 33. L. Hirst, and E. J. Miller, Zoc. cit70 ANNUAL REPORTS ON THE PROGRESS OF CHEMISTRY.acid (XXIII). Direct support is thus given to the butylene-oxidering €ormulz for these derivatives of y-glucose.Q01H*F---'MeO*Q*H QO2HMe 0 *Q *€€MeO-Q*H I --+ MeO*y*HCO,HI QH*OHH*Q-H*Q*OMe(XXIV.) (XXV.)CH2*OMe (XXVI.)Similarly, tetramethyl y-mannonolactone (XXV), which hasbeen ascertained to be structurally related to y-mannose-diacetone(XXIV), gives rise to a degradation product recognisable 33 asi-dimethoxysuccinic acid (XXVI). In no case is it observed that alactone from a y-sugar derivative yields on oxidation a trimethoxy-glutaric acid, and the above evidence permits of no suggestionthat an ethylene-oxide or propylene-oxide ring is present in thesey-sugars.Considerable importance is attached to the observations nowrecorded that d-trimethyl y-arabonolactone (XXIX) is intimatelyrelated t o and obtainable from tetramethyl y-fructose (XXVII).The latter sugar undergoes oxidation with nitric acid to give abutylene-oxidic 34 trimethyl lactol acid (XXVIII) which is degradedby acid permanganate t o crystalline d-trimethyl y -arabonol ac t one.The further oxidation of the latter confi~ms its previously deter-mined structure, since it has now been degraded 35 to Z-dimethoxy-succinic acid (XXX), and the crystalline methylamide of the latteris identical with that prepared from Z-tartaric acid.The enantio-morphic ,?-variety of trimethyl y-arabonolactone (XXXII), whichis derived from the y-form of trimethyl arabinose (XXXI), gave33 E. H. Goodyear and W. N. Haworth, loc. cit.34 J. Avery, W. N. Haworth, and E.L. Hirst, J . , 1927, 2308; A., 1057;W. N. Haworth, E. L. Hirst, and V. S. Nicholson, ibid., p. 1513; A., 859.s 6 W. N. Haworth, E. L. Hirst, and A. Learner, ibid., p. 2432 ; A., 1173ORGANIC CHEMISTRY .-PART I. 71on further oxidation the corresponding d-dimethoxysuccinic acid(XXXIII) , which was similarly identified.FH,*OMe _-__. 702H ----CH,*OMe CH,*OMe(XXVII. ) (XXVIII. ) (XXIX. ) (=x*)x-vo H*$J*OMe p 2 HMe0.Q.H + H*F*Ol\leMeO*(-i*HC0,H(XXXI.) (XXXII.) . (XXXIII.)--+L V * HCH,*OMe CH,*OMeThese experimental results appear to admit only of the interpret-ation that the y-lactones derived from methylated aldopentoseshad been correctly diagnosed and the butylene-oxide formulaapplied to tetramethyl y-fructose in the Report of last year isfinally confirmed.It is also clear that the butylene-oxide structureallocated to derivatives of y-arabinose is established, thus bringingall the known y-sugars within a comprehensive generalisation.Collateral evidence of the butylene-oxide character of tetra-methyl y-fructose (XXVII) is furnished by the unexpected observ-ations that this sugar passes with remarkable ease,36 in contactwith either dilute mineral acid or with acetic anhydride and sodiumacetate, into w-methoxymethylfurfural (XXXIV), which wasidentified through its oxime and semicarbazone as well as throughthe corresponding acid (XXXV). The latter was compared withauthentic crystalline o-methoxynzethylfurancarboxylic acid prc-pared from ordinary fructose and also from glucosamine by theFischer transformation of chitonic acid t o o-hydroxymethyl-f urfural.CH( OMe)*C(OH)*CH,*OMe CH:C=CHO CH :C*CO,HCH( OMe)*CH*CH,*OMe(XXVII .) (XXXIV.) (XXXV.)H :C*CH2*OMeA mechanism for these transformations is advanced in explanationof the probable consecutive changes leading to the furan com-pounds. In this connexion it is also of interest to notice thatw-hydroxymethylfufural is formed from pine lignin by steam.37I >oa6 W. N. Haworth, E. L. Hirst, and V. S. Nicholson, Zoc. cit.87 W. Fuchs, Ber., 1927, 60, 1131; A., 66072 ANNUAL REPORTS ON THE PROGRESS OF CHEMISTRY.Two crystalline derivatives of y -glucose have now been isolatedas the cy.- and p-forms of pentabenzoyl y-glucose.38 Their specificrotations ([.ID + 58.6" and -52.6") differ widely from those of thecorresponding normal glucose derivatives ( + 107.6" ; + 23.7").The latter values correspond fairly closely to those of a- and p-ghcose (+110" ; +17*5") and from these comparisons it may besurmised that, were it possible to isolate the two y-forms of thefree sugar, the a-form would have a much lower positive rotationthan the normal or usual form of a-glucose, whereas the p-formof y-glucose would be strongly lzvorotatory.Again, the aboverotations of a- and @-glucose are comparable in their range withthe values of the normal a- and @-methylglucosides, and it wouldappear that the two stereoisomeric forms of y-methylglucosidemay be expected to possess specific rotations similar in range tothe above pentabenzoyl derivatives of y-glucose, the p-form ofy-methylglucoside having a much higher laevorotation than thenormal p-methylglucoside.Tetramethyl 7-glucose is given an equilibrium rotation of [.ID - 11" and it is reported39 that the m.p. of one form is slightlyabove 0". It has again been shown that this sugar is formed in aseries of stages from glucose-diacetone, to which is thereforeattributed the butylene-oxide structure of a y-sugar. A viewexpressed in last year's Report, that in the formation of glucose-diacetone from crystallinc glucose the amylene-oxide ring of thelatter undergoes displacement, is supported by the authors of thesame paper.The use of the expression " normal sugar " was admitted to theterminology of the sugar group for the reason that it served todistinguish those sugar derivatives related to the " normal " a- andp-methylglucoaides from the so-called " y-sugars," which wererelated to the less stable y-methylglucoside isolated by Fischer in1914.Derivatives of sugars other than glucose have also beenreferred to the same two normal structural types, represented byCC- and 8-methylglucosides on the one hand and by y-methylglucosideon the other.The a- and p-methylarabinosides are normal in the sense thatboth are structurally related to a- and p-methylglucosides, and inthe same category are also p-methylxyloside, p-methylgalactoside,a-methylmannoside, p-methylfructoside, and a-methylrhamnoside.Thus the common forms of methylhexosides and pentosides are'' normal " and possess the amylene-oxide structure.Future workwill doubtless reveal crystalline methylhexosides and pentosides38 H. H. Sclzlubech and W. Huntenberg, Ber., 1927, 60, 1487; A., 858.39 F. Micheel and K. Hess, Annalen., 1926, 450, 21; A,, 1927, 43ORGANIC CHEMISTRY .-PART I. 73which conform to the butylene-oxide type of y-methylglucoside,which is at present recognised only as a liquid mixture of fwostereochemical forms.Knowledge of the structure of sugars has reached a stage atwhich, it is suggested,k0 confusing terms such as normal, y-, h-,amylene-oxide, butylene-oxide, 1 : &oxide, 1 : 4-oxide might con-veniently be replaced by a reformed nomenclature having a definiterelation to the structure and configuration of the sugars.It is seenthat the parent form of " normal '' or amylene-oxide sugars isrepresented by pyran, and of the y- or butylene-oxide augars byfuran,CH*OH CH*CH,-OH\ // CH*OH(A)and thus the reduced and hydroxylated parent forms represent thesimplest sugars. For example, (A) is an expression of the simplestnormal pentose, and (B) is that of the simplest tetrose-a y-sugar.(A) and (B) may conveniently be named pyranose and furanose.The simplest y-pentose would be (C), wherein a side chain is attachedto (B). The configuration of each sugar may be represented by theprefix xylo-, arabo-, etc., so that the spatial distribution of thehydroxyl groups, as well as the ring structure, would be clearlydefined by a terminology : xylo-pyranose, arabo-pyranose , etc.,and xylo -furanose, ara bo -f uranose, etc.(D)Similarly the two types of hexoses (D) and (E) may be termed gluco-pyranose and gluco-furanose, and the variation of the prefix tomanno-, galacto-, etc., would define the structure and configurationof the remaining aldohexoses.40 E.H. Goodyear and W. N. Haworth. Eoc. cit.c 74 ANNUAL REPORTS ON THE PROGRESS OF CHEMISTRY.HO.CB,~SOH / /*\ \ FH2 HO*CH,*fOH /O\ gHcrr,oHH-OH CH-OH E(0H)- HOK“ C L * (G)(F)The expressions fructo-pyranose (F) and fructo-furanose (G)correctly define the two constitutional forms of fructose.It must, however, be kept in view that both aldoses and ketosesmay, and probably do, react also as the open-chain aldehyde andketo-forms, and that this transitional phase is present in sugarsolutions which are undergoing or have undergone mutarotation.In the condition of equilibrium it is by no means impossible tocontemplate that the pyranose, furanose, and open-chain types ofsugars may all be present; but in the case of glucose, which hasbeen most widely studied, there seems little doubt that the pyranosetype preponderates even in aqueous solution.This problem isengaging the attention of physical chemists, and a clear statementof the present position of this line of inquiry has been contributed.The account of this given by T. M. Lowry 41 serves also to sum-marise the most recent results obtained in the study of the mutarot-ation of glucose and its derivatives in various solutions, and referenceshould be made to this interesting paper.(H) (K) (L)The synthesis of the natural glucoside coniferin is reported.42This has been effected by condensing acatobromoglucose with thepotassium derivative of 4-hydroxy-3-methoxycinnamaldehyde, andhydrolysis of the acetyl residues in the product.The synthesis of indican43 has been accomplished by the inter-action of rncthyl 3-hydroxyindole-2-carboxylate with acetobromo-glucose in acetone solution containing potassium hydroxide.41 2.physikal, Chem., 1927, 130, 125.42 H. Pauly and K. Feuerstein, Ber., 1927, 60, 1031 ; A, 649.43 A. Robertson, J., 1937, 1037ORGANIC CHEMISTRY .-PART I. 76Hydrolysis of the acetyl groups yielded 3-hydroxyindole-2-carb-oxylic acid- p-glucoside. Heating with sodium acetate and aceticanhydride led to the elimination of the carboxyl group and also tothe formation of an acetylated product, which was identical withpenta-acetylindican.The latter was deacetylated with methyl-alcoholic ammonia and gave indican, identical with the naturalglucoside.Two glucosides of alizarin have also been synthesised,44 butneither of these is found to be identical with ruberythric acid.Some progress has been made in redetermining the constitutionof the acetone-sugars, since the earlier formuh are of doubtfulvalidity. It is demonstrated that galactose-diacetone 45 is to berepresented by (H) (since it is convertible into fucose-diacetone) ;and a- and P-fructose-diacetones 46 by (K) and (L).Disa;ccharides.-The synthesis of disaccharides by simple anddirect union of hexoses constitutes a new and facile method.47It is reported that maltose is obtained by heating equal weights ofa- and P-glucose at 160" in a vacuum.Similarly, lactose is said tobe formed by heating p-glucose and P-galactose for 30 minutes a t150"/15 mm. in presence of zinc chloride, and melibiose, accordingto the same authors, is formed with almost equal facility.Galactose-diacetone condenses with acetobromoglucose withthe formation of a substituted di~accharide.4~ The free biose wasisolated as a dextrorotatory 6-glucosido-galactose. A synthesis ofthe naturally occurring primeverose has been accomplished bycombining49 1 : 2 : 3 : 4-tetra-acetyl glucose and acetohromoxyloseand hydrolysing the hepta-acetate to the disaccharide.Theconstitution of this sugar is therefore found to be (I) and it isprobably formed in nature from gentiobiose.The structural formula (11) deduced by earlier authors for maltose 50has been investigated by the progressive degradation of the biosethrough its oxime to a glucose-arabinose, which forms an osazone,and finally to a glucose-erythrose (111), which is found to be incapableof forming an 0sazone.~1 It is argued that the biose linking in thelatter must engage the second hydroxyl group in the erythrosechain, which is the equivalent of the fourth hydroxyl group in the44 E. Glaser and 0. Kahler, Ber., 1927, 60, 1349; A., 752.4 5 K. Freudenberg and K. Raschig, ibid., p. 1633; A,, 858.46 H. Ohle, ibid., p.1165; A., 649.4 7 A. Pictet and 11. Vogel, Compt. rend., 1927, 184, 1512; 185, 332; A.,4 8 K. Freudenberg, A. Noe, and E. Knopf, Ber., 1927, 60, 238; A., 230.49 B. Helferich and H. Rauch, Annalen, 1927, 455, 168; A., 859.50 W. N. Haworth and S. Peat, J., 1926, 3094; A., 1927, 135.5 1 G. Zemplhn, Ber., 1927, 60, 1555; A., 859.752, 960; Helu. Chim. Acta, 1927, 10, 280; A., 45076 A"UU REPORTS ON THE PROGRESS OF CHEMISTRY.reducing glucose residue in the original maltose. The formula (11)receives support from these conclusions, and also from the observ-ations of other authors who have studied the rate of lactone form-ation from maltobionic acid.52 The results indicate that the bioselinking is a t position 4 in the gluconic acid residue and that a six-ringlactone is formed by union of the acid group with the hydroxyl atposition 5.I(I.) HO*Q*HH*Y*OH 1CH2*OHHQ--'CH,*O*xylose CH,*OH(111.1 CH,*OHCellobiose (IV) is shown by the following observations to bebuilt up on the same structural plan as maltose.Oxidation of thebiose to cellobionic acid (V), followed by esterification and methyl-ation, yielded methyl octamethylcellobionate (VI) and by hydrolyticcleavage there were isolated the crystalline 2 : 3 : 4 : 6-tetramethylglucose, and also 2 : 3 : 5 : 6-tetramethyl gluconic acid (VII), whichgave on heating the corresponding crystalline 2 : 3 : 5 : 6-tetra-methyl y-ghmnolactone (VIII). This was identical with a specimenpreviously isolated from ( a ) methyl octamethylmaltobionate andCH,*OH (IV-1 *CHZ*OH CHZ*OH W.) *CH2*OHT02Me H- TO2H TO--1I (iH*OMe I QH*OH YH-YH2QH*OMe d?H*OMel QH-OMe yH*OMe 0TH-OMeJ yH*OMeo j YH-OMe --+ yH*OMeQHoOMe yH*OMeCH,-OMe CH,*OMe CH,*OMe CH,*OMe(;'H-OMe FH-WI.) (VII.) (VIII.)62 P. A. Lek-em and H. Sobotka, J . Biol. Chem., 1927, 71,471ORGANIC CHEMISTRY.-PART I. 77also from (b) 2 : 3 : 5 : 6-tetramethyl y-glucose, the structure ofwhich had already been determined by oxidation methods. Theconstitutional formula (IV) which had previously been determinedby other methods and by the sa,me authors 53 is thus confirmed.An analogous investigation was undertaken .to determine thestructure of lactose. Lactobionic acid gave on methylation methyloctamethyl-lactobionate, which on hydrolysis yielded crystalline2 : 3 : 4 : 6-tetramethyl galactose and 2 : 3 : 5 : 6-tetramethylgluconic acid (VII).The latter acid was transformed by heat intothe same crystalline lactone (VIII) as that which was isolated alsofrom (a) methyl octamethylmaltobionate, ( b ) methyl octamethyl-cellobionate, (c) 2 : 3 : 5 : 6-tetramethyl y-glucose. It follows thatlactobionic acid 54 has the same structural formula as cellobionicacid, modified, however, by having a galactose instead of a glucoseresidue a t the position marked * in the formula (V). The con-stitution allocated to lactose by previous authors is again confirmedby these data. The revision of the formula of glucose has servedadmirably the general plan for the formulation of the disaccharides ;and the determinations of constitution of lactose, cellobiose, gentio-biose, and melibiose during the pre-revision period fall naturallyinto the newer system of expression in which the constituent hexosesare shown as amylene-oxides or pyranoses.Maltose also conformsto this general plan.A disaccharide formula which has been affected by the revision ofthe older oxide-ring applied to the simple hexoses is that of sucrose.It is now shown that normal fructose is an amylene oxide or pyranose,and that the older butylene-oxidic formula assigned to this ketosemust now be allocated to the y-fructose, which is the componentoccurring in sucrose. It is thus established that the normal sugarswhich have been so far investigated are based on a common pyranosestructure, whether aldoses or ketoses, and y-fructose takes its place YH- VH,*OH[T€€*OH?-o-TFH*OH 7 4 (IX.) LYH-OH YHo0Y YH-CH,*OH YH CH,*OHalongside the y-aldoses as a butylene-oxide (or furanose) form, Theproofs leading to this conclusion are given in the section undermonosaccharide^,^^ but since the tetramethyl y-fructose isolatedfrom methylated sucrose has been utilised as the material on whichj3 W.N. Hamorth, C. W. Long, and J. H. G. Plant, J., 1927, 2809.54 W. N. Haworth and C. W. Long, ibid., p. 544; A,, 450.b5 See pp. 71, ‘7478 ANNUAL REPORTS ON TEE PROGRESS OF CHEMISTRY.the investigation is based, the structural formula of sucrose (IX)was naturally the main issue in these determinations, and thisformula, advanced last year, has received substantial confirmation.Whereas maltose, cellobiose and lactose are similarly constitutedin that (their essential stereochemical differences being neglected)they are all structurally represented by the formula (X),,/""\""\/' 1 ICH*OH /T 0CH*OH I CHCH,*OH I 1:::; :H*CH2\CH*OH \ /'QH 0 I CH CHCH*OH 0CH*OH CH*CH,*OHI I\ //\CH*OHI I6 1 ~ ~ 0 ~ CH*CH,-OH(XI.)the important bioses gentiobiose and melibiose conform to thestructural type (XI) wherein the non-reducing hexose component isattached t,o the reducing component through a hydroxyl linkedto the side chain of the latter, and not to a hydroxyl attached to thering as in (X).These formula for gentiobiose and melibiose were advanced someyears ago.The constitution of gentiobiose was verified last yearby synthesis, and this structure for melibiose is strongly supportedby two further papers.56 Methylated melibiose (XII) gave onhydrolytic cleavage the crystalline 2 : 3 : 4 : 6-tetramethyl galactose(XIII) and also 2 : 3 : 4-trimethyl glucose (XIV), recognised as itscrya talline glucosid e.QH*ORle ryH-rQH*OMe I YH*OMe1*QH*OMe J YH-OMe * 1 CH-OMe io FH*OMe I ---+ H*OMeCH,*OMe7H-JCH,--' CH2*OMe CH,*OH6e W. Charlton, W. N. Haworth, and W. J. Hickinbottom, J., 1927, 1527;(XII.) (XIV.) (XIII.)A., 859; W. N. Haworth, J. V. Loach, and C. W. Long, ibid., p. 3146ORGANIC CHEMISTRY .-PART I. 79The union of these two hydrolytic fragments could only haveexisted through the exposed hydroxyl in the terminal posittion in(XIV), from which it follows that methylated melibiose has thestructure (XII) and the free diaaccharide the corresponding formula,indicated also in another way in (XI).Again, melibiose gave onoxidation with bromine melibionic acid (XV) ; and this structuralformula alone could be applied to it, since the fully methylatedmelibionic ester (XVI) gave rise on hydrolytic cleavage to crystallinetetramethyl galactose (XIII) and to tetramethyl gluconic acid(XVII) which, on further oxidation, was transformed into thetetramethyl saccharic acid (XVIII).CH2-J CH,*OMe(XVI.) II IYH*OMe (XVII.)CHaOMe(XVIII.) QH*OMeCH*OMe$H*OMe $H*OMeCO,H CH,*OHThis structural formula for melibiose has been attacked byG.Zem~I&n,~7 who, by degrading the sugar through its oxime,failed to isolate either the galacto-arabjnose or its crystallinephenylosazone. The author attributes this failure to the presenceof the biose linking in the position shown below in the galacto-arabinose (XIX), since this would prevent osazone formation ;and the formula suggested by the same author for melibiose is (XX).[YH-O-’ YH&l r?H*OH ~~~~1 (xIx.) 0 YH*OH QH*OH b 0 (iH-0 QH*OH 0 (xx,) LYH~OH QH*OH I LTH-OH THOOH ITHOOH r Q H YH-OHCH2 FH---] TH QH--’CH,*OH CH2*OH CH,*OHAn alternative reason for the difficulty experienced by the authorin isolating the expected products may well be that on the basisof the melibiose formula (XI) which he attacks, the galacto-arabinoseb7 Ber., 1927, 60, 923; A,, 54580 ANNUAL REPORTS ON THE PROGRESS OF CHEMISTRY.obtained by degradation would be a butylene-oxide or y-sugar(XXI), and hitherto no crystalline osazone of this labile type hasbeen isolated.CH,*OHThe synthesis of a new disaccharide has been effected as follows :1 : 2 : 3 : 4-tetra-acetyl glucose was condensed with acetobromo-galactose in presence of silver oxide.The resulting octa-acetateof the glucose- p-galactoside (XXII) was not identical with melibioseocta-acetate, nor was the liberated disaccharide the same asmelibiose.A TH*OAc A TH*OAcFH*OAc QH*OAc 1V$X&*OAc 4- 1 QH-OAcryH-OAc rQHBrCH,*OH CH,*OAc CH,---’ CH,*OAcThese authors 58 rightly conclude that either the configurationof melibiose as a P-galactoside is incorrect or the Btructure pre-viously ascribed to this biose (XI) is untenable.The configur-ation given to melibiose in the literature is based on the observation s9that this sugar is hydrolysed slowly by emulsin. Other authors 6oare unable to accept the interpretation which has been given tothis evidence based on enzyme cleavage, and support the argumentthat melibiose, having the recognised structure (XI), cannot bea glucose-p-galactoside, since on this assumption its high rotationis in disagreement with Hudson’s rule and with other data. Onthe view that melibiose is a glucose-a-galactoside, correspondingwith maltose as glucose-a-glucoside, the newly observed facts canbe reconciled.The constitutional formula ascribed to turanose and reportedlast year has received experimental support from another worker,who outlines a, much more convincing proof of the character ofthe trimethyl y-fructose component isolated from methylated(XXII. )63 B.Helferich and H. Rauch, Ber., 1926, 59, 2655; A., 1927, 44.50 E. Fischer and E. F. Armstrong, Bey., 1902, 35, 3144.6o W. Charlton, W. N. Haworth, and W. J. Hickinbottom, loc. cit.; W. NHaworth, J. V. Loach, and C. W. Long, Zoc. citORGANIC CHEMISTRY .-PART I. 81melezi tose. The constitutiona1 formula for the trisaccharidemelezitose appears, therefore, to be established.Gl There are, how-ever, difficulties in interpreting the behaviour of enzymes towardsthis sugar, in that a preparation from Aspergillus niger, which actsslowly on sucrose, attacks melezitose rapidly, and one authorsuggests that the sucrose residue does not occur in melezitose.A new trisaccharide has been synthesised by condensing 1 : 2 : 3 : 4-tetra-acetyl glucose with acetobromocellobiose in presence of silveroxide.The trisaccharide 62 was isolated as a crystalline substanceand gave a crystalline omzone. It does not appear to be identicalwith a natural product.PoZysaccharides.-There seems little prospect of the cessation ofinterest in the allocation of new structural formulz to the chiefmembers of the group of polysaccharides. The volume of experi-mental work contributed on this subject grows apace, and if ageneralisation on them be expected of the Reporter, it must besaid that the more recent results have proved the futility o€ allattempts to express the formulation of cellulose or starch until ourknowledge of many underlying problems has outgrown its presentlimitations.Determinations of molecular weights of derivativesof the polyaaccharides lead in many cases to precarious reasoning,particularly when, as is too often the case, the products examinedare amorphous mixtures. The contest is proceeding betweenrival theories concerning the factors which are operative in thecomplex aggregations in polysaccharides.One view of the constitution of polysaccharides postulates theexistence of a comparatively simple structural unit 63 which retainsits definite entity while existing in a highly associated form. Underselected conditions there may be a lesser degree of associationoperative in certain solvents, which means that the solute undergoesdispersion.Presumably the linkages responsible for this associationare not those of ordinary valency. The only alternative seems tobe to regard these as being due to the operation of co-ordinateco-valencies on the model of N. V. Sidgwick's suggestions.64 Thisrenders the formula originally suggested by K. Hess a very attractiveone,e1 (Miss) G. C. Leitch, J . , 1927, 558; A,, 450; M. Bride1 and C. Aagaard,62 B. Helferich and W. Schafer, Aianalen, 1926, 450, 229 ; A., 1927, 135.63 M. Bergmann, Annalen, 1927, 452, 121; A., 341. Compare Ann.Cornpt. rmd., 1927, 185, 147; A., 859.Reports, 1924, 21, 91.Presidential Address to Section B, Brit. Assoc. Reports, 192782 ANNUAL REPORTS ON THE PROGRESS OF CHEMISTRY.It is difficult, however, to conceive of the persistence of theseco-ordinate links when the hydroxyl groups of cellulose have beenreplaced by methoxyl or acetoxyl groups, and if for no other reason,it would appear that this consideration renders the formula of Hessdifficult of acceptance.On another plan, the simple structural unit is held to polymeriseto a higher unit or to higher units of varying complexity by struc-tural change.65 The breakdown of these complexes is then a step-wise process ending with the simplest possible form, which canagain revert to the more complex state. In this connexion, theilluminating example of Z-trimethyl 6-arabonolactone may becited. This crystalline substance polymerises with extremereadiness to a polymeric crystalline form, which has ten times themolecular weight of the original lactone and reverts to the simplelactone on heating. There seems to be on the whole little differencein principle between this second view of a polymerised unit and thetraditional view which adopts the conception of hexose rings linkedby connecting oxygen atoms in a, chain of indefinite length by theoperation of ordinary valency. It does not seem to be an essentialconsideration that the unit which undergoes polymerisation shouldretain, on this second plan, its structure in the polymerised aggregate,nor does it necessarily follow that the process of depolymerisationto the simplest unit would always be attainable. A modificationof the older, traditional formula has been advanced and has beenbased on a study of the X-ray spectrograph of cellulose in ramiefibre. The linkings between the constituent hexose rings, accordingto this formula, engage in alternative pairs the 1 : 1 -hydroxyl groupsand the 4 : 4-hydroxyl groups.66 It is also suggested that lateralcontact with hydroxyl groups of adjoining chains establishes afurther type of union by the functioning of subsidiary valency forces(co-ordinate valencies) which are more readily dissipated orovercome. etc. etc.