THE CONSTITUTION OF POLYSACCHARIDES. PART 11. 1459 CLXV.-The Constitution of Polysaccharides. Part 11. The Conversion of Cellulose into Glucose. By JAMES COLQUHOUN IRVINE and CHARLES WILLIAM SOUTAR. IN the seriw of investigations on the constitution of pollysaaahasides with which we are elngaged in tlhis laboratory a prominent pla,w is naturally assigned to the more! definite varieties of cellulosel. The original literature on the reactions and colnstitutioln of celluloae is voluminous but it aa.nnoh be claimed that views regarding even tlhe fundamental nature of the omplex are by any means establisheld. This obscurity is not surprising considering the special difficulties which surround constitutional studies of this type. Even if wllu-lose aan be regarded 8,s a chemical individual in t'hs ordinary sew 1490 LRVINE AND SOUTAR: t3he customary methods of sohh,g problems of structure are of little ava.il in view of the insolubility of the oompound the dubiety attending its molemla,r magnitude its behaviour as a.codloid a,nd t>he probability tbatl a fibrous structure is not chmica,lly homo+ geneous. As a result ma,ny of t)he st'at'emenb which find a place in tlhe permanentl litetmture a,re bamd on very insecure evidence, and a(re even contradictory. In reviewing brielfly the premnt posi-tion of the subject? reference may at tlhis stage be limitled to recen.t papers whioh have a belaring on fundament,aJ questions. A normad cofton cdlulose is represented by the formula (C6HlOO& an.d idem a,s t'o the molecular st>ruoture ham been formed largely oan the evidenm afforded by hydrolysis.It is impor-tarntl to1 note t,hatl in this particular case hydrolysis is not rea,dily effected and invollves the use of somewhat drastJc relagent.s so thah baking into aacount the unsta.ble nature of the hydrolysis p r d u d e , semn.da.ry rea,ctdons are inevitable[. The fact that numerous and complelx degra8dation colmpoands are formed has thus given rise1 t'o conflictsing opinions on the chemical nature of celluloge butl itl is unnecessary in the present pape:r to discuss theset rival theories in detlail. Appareat,ly tlhe view which finds most acueptanm is that) cellulose, like starch is essentidly a polyglucoae anhydride and it may be weill to state at once that this conception admi.ts of a double inter-preta,tion; The complex may consist of the simple units C,H,,O, (derived from a hexose by molmular loss of wa,te;r) pdymerisd in unknown numbers.On the othelr hand n molecules of as hexose may be direatly connested together through the elimin.ation of molsmles olf Walter or where n is a la.rge fa8ctlor of m- 1 molemulee. I n either ca,se t,he first point which must be seltltled is to ascertain bey0n.d doubt if glucolss is a.atua,lly tlhe hexme folrmed by the hydro!-lysb oif cslluloue and if sol t.o determine exactly the amountl of sugar tlhus produmd. The inquiry becolmea muah more definite if it can be shown tlha,tl, witahin the1 limits of reasonable experiment(a1 error cellulose can be converted in.tol gluc;ms in terms of the equafdoa : (CGH1O05)n f nH,O * mC6H,,O6* 100 parrtrt.s + 111.11 pasts.From timet to time wnfidelnt statements appear in tjhe literature that practliaaJly quantitIative yields of glucose have been obta.ined from oellulm but the grounds upon which suah claims ar0 made are by no means convincing to workers in the sugar group. Praoti-cally speaking the only experimentad methods available for degrad-ing ~lluloae depend upm the use of mineral acids either alon THE CONSTITUTION OF POLYSACCHARIDES. PART II. 1491 or in colnjundlion with a,wtic anhydride and it is evident that any sugar t'hus liberated must undergo profound alteratioln when kept in contad wit,h these reagents. This no doubt) a,ocounts f o r the fact that hitheirto pure crystdine glucose has nevelr been obtained from t,his polys,accha.ride.Nevertheless Flechsig (Zeitsch. physioZ. Chem., 1883 7 523) claimed that a yield of 95-98 per cent. of the thee r e t i d ammint of glucose wits fmeld by t)he adim of sulphuric acid on cellulose but the stlatmelnt i,s based solely on the reducing power of a complex mixtlure and has little significance. Sahwalbe and Schultz (Ber. 1910 43 913) supplemented Flechsig's experiments by isolating the produo& of hydrolysis and obtained a se,mi-oryst,al-line suga,r amounting only to 20 per cent. of the tlheoretlical yield. Working on similar lines Ost atnd Wilkelning (Chem. Zeit. 1910, 34 461) ma,de t,he importmantl claim tha,tl the yield of glucose was a.lmostl quantitative but it may be remaskeld t$ha,t they exa,mined the1 produds otf hydrolysis po,lasimet,rically and reported specific rotatboas ra.nging from + 29'4O to + 44'8O whea-eas the equilibrium value for glucose is + 5 2 * 5 O .As a furtlher mea,ns of estima,ting the amount of sugar farmed tahq adopteld melthods depending on the reduction of copper so11utioq but in this case a h irregularit*iee were experienced and the results indicated yields of glucose varying from 73.4 to 113.5 per centl. of the welight of cellulose tlreated. The we of hydrochloric a'uid for degrading cellulose is even less satisfactory. As is well known Willstiitter and Zwhmeister (Ber., 1913 46 2401) dissolved coltltan-wooll in 40-41 per cent;. aqueous hydrolchloric acid and allowed the hydrolysis to protoeled in the cold. 'The oouw of the rea.dion wats followed polarimetlrically and, by data obtained from control experiments in which glucose was dissolved in the same acid medium they ca410ula.ted tha't the yield od tIhe hecmse formed from cellulose a,moiunted bo 96.3 per cent!.of the theoretical value. This mnclusion was appa8reatlly colnfirmed by the results of titrations which indicated the forma.tion oif 94 per ce8nt4. of the theoretical weight of glucme. The results appmr mn-vincing untJl th&y are wnsidered in mnjunct'ion with the known effects of hydrochloric acid upon glucose. %illstlatter and Zech-melister were of the opinion tha,t, owing to the l m mnmntration of sugar in the aoid mlutim no isomalbe was f o m d butl it has been shown (Davis J . SOC. Dyers m d Col. 1914 30 249) tlhat the action of hydrochloric acid in prolmating the antmcoindeasamtlion of glucose extbnds t,a solutions containing as lit'tle as 1 per mat.of the sugar. In addition few reagenta effect more fundamentad changes in reduhg hexoees than hydrochloric acid in either dilute or con-centrated solution. Thus traces of the acid convert the butylene-oxide forms of glucose into the ethylene-oxide isomerides. I n highe 1492 IRVINE AND SOTJTAR: concentrations of acid complex changes are therefore to be expected, 501 that when glucose is dissolved in 44.5 per cent,. hydrochloric acid t.hel specific rotation is + 164*6O a.nd thus exceeds the maximum value for a-gluc!osel by a,pproxima,t,elly 50°. Tha,t Willstatqtelr's prolcew has velry lithlel bearing on the prima,ry constitution of celllulolse is further showfi (Cunningham T.1918 113 173) by t,hel famothat both coft't80n and elspasto mlluloms give praatically idelntioal rot'a,tion curves wheln hydrolysed with coaceatsatleid hydrochloric a,cid undelr t,he colnditiolns described by him. I n fact the evidence of speufifio rot,atioln and reducing polwelr emn when apparently consistentl, caa,nnolt be held to chara.at,eirisel an uncrysta~llisable~ syrup as a definite sugar. A colnsidelrablel a,dvance oln ths use ob minera,l a,cids is marked by the; coinveirsioln olf mllulose~ into glucose a,cet,at es as elaborated in the exha;us;tiive reseasches of Ost and his pupils. Recopisable crystal-h e products coaslisting of cedlobiolsei oIct,a,-aceltatee a.nd glucolsel pentam-acetates are thus obtained ro that i t is possible to ascribe t,rustr worthy va,lues to the yields.By using amltic anhydride1 conta,ining a8pprolxima.tdy 10 per centl. od sulphuric acid as tlhe hydrollytio reagent Ostl (Chem. Zeit. 1912 36 1099) i,sola.ted a mixture off solid acetates amounting to 60.6 per cent,. of the1 theoretical quan-tity the rema.ining prolducta being uncrystallisable syrups. It. is very doubtful if the latter can be inoluded in cra3culatJng the! t'oltal yielld oc if poda,rimelt(ric mehholds a8re a8dmi.ssible in wtimating hexose aue,t,atee as many factors comb,inei ta render suoh a method uncer-tain (Hudson and Pa,rkm J . Amer. Chem. SOC. 1915 37 1589; Hudson ibid. 1591; Hudson and Johnson ihid. 1916 38 1223). Up to1 the1 preiwnt time in no reaelasch on t'he hydrolysis oif cellu-lorn where] a yield of glucose even approlxima,ti.ng tol the theoretical amount has been claimed ha,ve tlhe results been based oln t'hei quan-tity of the s;ugar or of a crharackeristia derivahivel amdually isolabd.In the1 work nolw described we0 a.dhere1d to the prinuple tlhatl the yield of hexose should be ascert.ained from the weight of crystalline mmpolunds obtained in a mndi,tioln of a.nadytlical purity and in well-defined sterelolahemiclad forms. Adolpting tlhis standard we have been able t,ol sholw t,hat as a minimum the yielld od glucose obt<ained from cellulose is 85 per cent. olf the theloreitid amoant. The melt,hold used by us elmbodies the same principle as amtolysis in that it involved hydrolysis of cellulose and simultaneous con-densation of the sugar liberated so as to give a &able derivafive which t8helre.a9t,er relmained una.ff eldeld.I n this way the1 glucose wa's proteobd from the destructive effect of the hydrolytic agenk. The material employed was a normal cot,tm celluloee for a supply of which we are indebted to Mr. Wm. Rintoul of Nobel's Explosive THE CONSTITUTION O F POLYSACCHARIDES. PART 11. 1493 Co. This was treated as described in the experimental part, with a 1a.rge excetw of acelt4ic a,nhydride containing acelttia aad snlphuric wids. Wheln thel fibrous stlructare ha,d diaappelare:d tlhet produatl was poured intlol Walter and the precipitlatetd solid ,setpasated. The filt,ratle t'hen cont'aineld t'hs lower acetyla.te,d glucwes t,oIgelther with amtosulpha,tea and other soluble degrada,tion productx whilst the insoluble1 residue1 colnskteld of polysa,cahafidel ammtlattes.On heating the la,tt,elr in an a.utoclavve a,t looo with mstlhyl alcohol1 aolntaining 0.5 pelr cent'. of dry hydrogen chloride the first e$e& was ts remolve the acetyl groups which were converted into methyl acetate (Perkin, T. 1905 87 107; Fentoa and Beirry Proc. Camb. Phd. Soc., 1920 20 1 16). Thelreaf tsr simultlaneoas hydrolysis and coln-densat,ion witlh the sollvetntI ensued the process thea being pa,rallel witlh tlhe conversion of stlarch iatol metlhylglucoside (Pisoher Ber., 1895 28 1151). The main productl of the relaction consisted of arystlalliae methylglucolsideA but a,boutl 25 per cent. od t.he maltelrial persistled aad remained pra8ctbcally unafiected on relpelati.ng tlhe ttrea,tlment4 with the a.cid alcohol.This anmrphous relsidue was t8herefolre hydrollysed by melans olf dilute1 aqueous hydrochloric acid, and the product aga,ip brolught into rea,dio'n witlh acid methyl a.lcoho1. In this way the tot.a.1 yield of metIhylglucoside froim the fra.&ioln insoluble in water was ascertained. Owing to the large volumes which had to be manipulated the t8relat8ment od the products soluble in water was laborious. After removal of t4he free a,cids t4he solveint was eva.polra8tle.d a'nd the s u p converted into mettIhylglucoside in the usua.1 way. In every case bhhs rnet8hylgluemide wa,s cYbt,aJned as a co~lourless syrup which rapidly solidifid tol a hasd mass of cayst.als. On hydrolysis wit,h acid no difficulty was experietnced in obt.aining pure crystalline glucoael frolm the1 glucoside.Stutement 01 Yields Obtained. Two1 melt4hds ob estimating the yields of melt,hylglucosidet wetre employed. The weights reported refer tlol pure crystallinel material drietd in a vaauum weln until comtlantj and as an additiolnal ahecrk, the olptiaal activity ob eaah sollution whioh yielded methylglucmide was datelrmineld. During the glucolsidel folrmattioin heating with acid mehhyl alcohol was coatinued untlil equilibrium had belen esta(b1ishetd between tlhef a- and P-folrms and under the conditions adopted the propmtboln in which kheae modifications are present is respectively a = 77 fi = 23 per cent. (Jungius Proc. K . Akacl. 1T'efemch. Amsterdam 1903 6 99). The mnstantI spedfic roltatioii attained when sollutims of melthylglucwidel in acid methyl alcoho 1494 IRVINE AND SOUTAR: a're heatled is thus of the ordea of + 1 1 4 O so that we w m able t,o check our gravimetlriria r e d & pdarimetriaally.As we rely exclu-sively oa our gravimetlri.a data, this prelca!utlioln may swm unneces-sary butl it8 a,doptioa selaured t,haat expelrinietntal conditions were weld whiah predudd a.ny possibility off the yields being af€elct,ed by t,he fo;rma,tioa sf y-metlhylglucoaide. The combined rtzsults of one typical experiment are shown in the' follolwing chart the weight of celllulosel t,a.keln (70 grams) being cotrrected for tlhe moistlure and ash content,. 112.0141 graiiis f i solid acetylated \-prod u c t . ,/ \ \ ,/" x I 3fe thylglucoside.I Cellulose J. 34.0513 grams 65.0614 grams 29.0830 grams -+ 25.9639 ,, resistant deacetylated 6.2340 ,, prod uc t . fl- -- - -\ /' 66 2493 grams = Total \ '., li i \ 10 litres aqueous acid solution. Expressing t'hel resultl in perwnta>ges : Cellulose -+ M e thylglucoside -+ Glucose 100 parts gave 101.8257 parts equivalent to 94.4775 parts. If the celluloael mollelmlel is composed entirely of glucose residues, 100 parts oif the podysacaharide should give 111.11 pasts olf glucose, so tihast the1 opeiration olf the above scheme gives a yield od methyl-glucmider (and thereforel od glucosel) of 85.03 per cent. of t>hs theof-retical amountl. Although the1 manipulations were conducted with a standard of accuracy comparable with that employed in gravi-metric analysis itl is obvioas that the above result is a minimum value.The preparation of methylglumside from glumel although a smooth reiaotion is not quantitlatlive owing to inevitable experi-mental 101% in isolating and crystallising the product. We do not propose however to introduce any correction in whiah allowmce is made for expwimental law a d our future work will inolude an attempt to alcwunt2 for this diveirgeizce of 15 per mnt. from the thesretiical value THE CONSTITUTION OF POLYSACCHARIDES. PART 11. 1495 Discwsion of Results. Althofugh the1 'main objectl of the presentl research did not involve the deltadd Colnstitution of wllulmel solme of the results obtained have1 a. dire& bearing on tlhe problem and may be disoussed. As a rule i t is an unoommo1n experielnm in the sugar group to1 obtain crystlalline derivatives in yields which exceed 80 per cent.and the figurw now submitked thus afford strong evidence that cotton oelllu-lose is composed essentially olf glumel rmidues condensed together. It is to1 be noted however thatl the processes adopted by us would noit serve to isolate any ketwe constituent should such be preisentl. I n view of the fact that cellulose yields bromomethylfurfuraldehyde (Fenton and Gostling T. 1901 79 361) itt is conceivable that the unexplained margin of 15 per celnt. may be amounted for by the presence of a nucleius in the cellulose molemle which is resolved into a keltmel on hydrolysis. It is possiblel although somewhat improb-able that the hexme units in mllulose are symmetricadly disposed as in inulin (Irvine and Steele this vol.p. 1474) and the alterna-tlive has to be kept in view tlhat two three or four hydrolxyl groups of individual glucose molecules may be involved in the coupling. Should tlhis be the1 oae the hydrolxyl content od cedlulose may still be regarded as three but this would be an average value and would not imply thatl as in inulin three hydroxyl grolups are present in eiery C unit. Evidelnce in support of this secolnd vielw has been colntsibutleld by Delnham and Woodhoasel (T. 1917 111 244) from the study of trimethyl cellulose and we hope in consultation with these uwkers to elxtlend tlhe investigation on melthylated wlluloses. Further evidence of the non-uniformity of the glucose linkages in cellulose1 is aflordad by the remarkable variatioln in the ease with which the oomponent parts of the modecrule undergo acid hydro-lysis.Another signifioanto faotolr is tlhatl the1 yield of cellobiose obtained from cellulose although varying greatly with the condi-tioiis of hydrolysis has never exceeded the maximum quoted by Klein (Zeitsch. cmgezo. Chem. 1912 25 1409). His figures are: Cellulose + Cellobiose octa-acetate -+ Glucose 100 parts gave 60 parts equivalent to 31.9 parts. From thme rmult,s it would appear that at least one-third of the cellulmel molecule mntains the linkage characteristic of celloibime. Now cellobiose contains eight hydroxyl groups one of which is a reducing group and therefore terminal. It follows that one of the reimaining hydroivy1 groups of the reduuing o p o n e n t must be attached to t4he reducing group of the1 second glucose residue.9310 position of this linkage may be fixd from among other factors 1496 IRVINE AND SOUTAR: tJhe constdtlutlion assigne.d to trimet,hyl glucose and f o'rmula I may t'hus bet deduced for oelllobiose : .CH*OH OH-0-G 'AH C!lH*O*CH.[CH*OH],*CH*CH(OH)*CK,~OH / \ UH-0-GI I I CH,*OR bB,*OH The! expanded stlrudare 11 in whiah G aad G relpresent glucose raidurn aan thus be delduced for a fragment olf the celllulose mm-plelx. Of the tIwo groups G and G, the latter i& the molre stable to hydrollyste and the systelm indicated in formula I1 evidentlly relpre-seats the mostl resistant portion of the cellulose mollede. It is significant that in olur wolrk we! encountelred tlhe same prolgrmive diaculty in eliminating the glucose residues f r m cellulow.The melthylglumside obtaineld was isolated from three1 groups of aaetollysis prolduc;ts : (1.1 (11.) A. Soluble in water. B. Insoluble in water and hydrolysed by acid melthyl dcmhol. C. Insoluble in water and resista.ntl to acid methyl dmhol. Our rwultsl tlhus show that tlhe cmlluloaa modeloule may be1 dis-sected into three1 portions and the a,pproximate ratliol in which the grcups A B and C are presentl is displayed below. Methylglucoside. Glucose. C,H,,05. cellulose /"-4 gave 9.5817 equivalent to 8.9 equivalent to 8-0 100 parts <+ $ 9 52.3372 9 7 48-6 Y Y 43.7 , 39.9068 99 37.0 9 9 33.3 - -94.6 85.0 It) will be1 wen tha,tl tlhe prolposrtioln olf C agreea approlxirna-tedy with the figure indicated by the maximum yield of cellolbiow oota-acet'ate obtlained from cellulose and it is our intlentlioliz to continue the investigation by tracing the structural distinction between the units A B and C.In addittion we hope to ascertain whether the glucose belongs to the ethylene-oxide or butylene-oxide types as the fundamental difference between cellulose and starch may depend on the nature of the oxidic linkage in the constituent hmose residues. E X P E R I M E N T A L . The material employed wit8 a normal mtlton cellulosel suppliefd by So the ReBearch Department of Nobel's Explosives Co. Ardeer THE CONSTITUTION OF POLYSACCHARIDES. PART 11. 1497 f a,r as the1 ultima8tet yields of met~hylglucoside are m~n.mtrned simila'r reaulte were1 obtaineld with filter paper which ww more resdily disintograt(d by the relagents.I n quoting yields allolwa,nce has been made f o r the moktarel content (6.7 per cent.) and the1 ash lelft on ignition (0.355 per cent.). Seventy g r a m o€ the1 mllulow cutq inta mad1 pieces 2 cm. by 1 cm. were placed in an enamelled iron beaker surrounded by a ba,t.h cooled wit,h running watelr. A mixture1 of 350 C.C. of amtic a>nhydride cont,aining 6-25 per celnt'. oC a.celtic a,cid a'nd 20 C.C. od conceatrated sulphuric acid was mo;leld to 1 5 O aad qui.ckly a,ddeld tlhe mass being rapidly agita,t,ed with a polwerful stirrer. The best retsults were obtadned when the maximum telmperatnre did not e l x d 75O and in twelnt~ minut'es the fluid mixture was poured into imcold water with continuous stirring during the dilution.After twenty-four hours the white precipi,tafe which had setltled became brittdel and was filtelreid a,nd washed until free from a.cid. The t,olttatl volume 04 iilt'rafe and wa.shings was 10 lit'ree a.nd tlhe insoluble aaeltlates after drying ah 40-50°/25 mm. weigheld 112 grams. The1 mat'eirial was sepa,ra,tled into tlhrea portlions a,coording to1 the1 solubility in. 95 pelr cent,. aJco.holl (1) soluble1 in the cold (2) soluble! only a t the boiling point, (3) insoluble. As each f ra,ctioa was mave1rt:ible into1 methyl-glu cmidel tlhis sepasa.tion was not' oarrield o u t in t'he 1a.rget-scale elxpelrim elnta. The following is a,n account of a tlypiaal experimelnt. Simultaneous Deacetylatz'on Hydrolysis and Xeth,ytation of t h e Inso Zub I e A c e t a t es.The mixed a.ceta,tes were disolveld in methyl alcohol containing 0.5 per cent. of hydrogen chloride m as to give a 5 per cantl. solu-tion and heated a t looo for seventy hours. On opening the auto-clave tlhe odour off methyl acetate was noted a8nd a white amor-phous precipitate was found to have collledd. This was filtered (Filtrate A); washed and d r i d (Residua B). Examination of Residue B.-The materia,l a?mounting fo 29 grams was a white amorphous powdelr insoluble in water, chloroform ether alcohol or dilute! hydroohloria acid but readily soluble in dilute sodium hydroxide solution. It melted and decom-pwed a t 237O and reached as a glucoside towards Fehling's mlu-tion beling hydrolysed on boiling with dilute acid.No chlorine was present and a mekholxyl determinatim gave a blank r e d t (Found, C = 44.0 ; H = 6-24 ; OMe= 0. C,H,,O requires C = 44.4 ; H= 6.20 ; OMe=O pelr cent.) 1498 IRVINE AND SOUTAR: Con.version of Besidue €3 into Methtylglucoside. On boiling under a condenser with 3.75 per cent. aqumus hydrm chlo,rio a'cid the white solid gra8duadly dissolved witlh the exwpt,ion of a small residue which in a. separate experiment was hydrolysed by means od 8 per cent. a$cid. Wheln the activity of the solutions was colmtant indiaatdng %hat hydrolysis was complete the liquids were unitled neutlradised wit'h barium ca,rbonate and evaporatd to drynegs under diminished pressure. A pale brown syrup remain&, which was extsa.&d five times with boiling methyl alcohol1 contain-ing 0.5 per cent.of hydrogen ahloride and the solution was heated aft looo untJl the1 optica.1 a<otivity remained const8ant ([aID + logo, cadculattleld oln the1 weightl of glucoside ischted). The amaid was then neutlraliseld by shaking the solution sucawsively with 1ela.d m d silver carbmmtes a.ftees which the filtrate wm boiled for some hours with cha,rmsl. A colourlew solution was thus obttained whioh 0.n concelntrafioln undelr diminished pressure gave a clear syrup. T'his3 rapidly crystallised t o a compact hard mass on the addition of a nucleus oif a-methylglucoside. The cryst'als were extracted five times with a large excess otf boiling ethyl acetate and the pure glucolside isolaf,ed in one crop from the united liquors.The product colnsisted otf t,he equilibrium mixture ot a- and P-methylglucosides. Yield, 25.9 grams. (Found C = 43-20 ; H = 7.41 ; ORXa = 15.31. Cak., C=43*30; H=7.22; OMe=16-98 per cent.) The specific roltatioa in wa.t'er (mean of t'wo detmmina.tdons) was + 114.B0 in place of the calcula,tled value + 114.0°. The compound behaved sha.rply as a glucoside towards Fehling's solution a,nd showed tche usual range of melting point (125-154O) for the mixed glu cornsides. Examination of Filtrate A .-This solutdon contained the equili-brium mixture8 of a- aad &methylglucosides in acid melthyl alaohol and showe'd [aIn + 113'8O (calculadad o'n the weight od glucoside obt,ained). It was neut,ra.lised as described above with lead and silver ca,rbonates the further treatmentl involved in removing CQ~-loidal silvelr aad in isolating the products being also identkal.As beforei no1 diffioulty was experienced in obtlaining the methylglum-sides in the crystalline condition. The product was free from halogen o'r sulphur and had no adtion on Fehling's sdutJoln untlil hydrollysed (m. p. 123-162O) (Found C = 43-30 ; H =7*15 ; OMe= 15.56. Calc. C=43'30; H=7.22; OMe=15.98 per oeatl.). The constant weight of pure glumside was determined by hea.ting atl 50° under diminisheld presure a currentl ogf air dried over phos-phori.a orxide being led tlhroagh t'he apparatus. A trap containin THE CONSTITUTION OF POLYSACCHARIDES. PART 11. 1499 tlhe sa,me dehydrahing agemt was placed between the receiver and the water-pump.Yield 34 grams. Preparation of Methylglucoside from 4 cetolysis Products Soluble in TVater. The aquelous acid filtxratle obtaineid in removing the insoluble acet,ates was distinctly dextlrmo8tatolry a.nd amoantled with wash-ings t.a 10 lit'res. This was subjeated to distilla,tion in &earn the voilume being kept constant, to remove the excess of amtic a8cid, a,nd the sulphuria a,ud present waa then precipitla8t.ed by shaking with barium carbonate. The filt,e,red solution was evaporated to dryness unde,r diminished pressure. A white crgsta.lline reaidue remaineld tsgethar with a yellow syrup which was elxtracted five times with boiling meithyl alcohol containing 0-5 per cent. of hydro-gen chloride( the elxtmction process extending over ten hours. The unit,ed extracts (Solution C) were separated from the semi-arystal-line undisso2lved solid (Rwiduei D).Examination of Residue D .-Alt(haugh largely inorganic Ohis residue contlaineld some orgaaia ma.tt8er delrived eiit,he,r from soluble acetmulphatels or from collolidad cellulose1 aceltatea. It was accord-ingly tlrelateld in exa'ctly the1 same mannelr as Residue B and the hydrolysis proldud isoIla.ted as a syrup. This was eixtaact\ed six tJmea with boiling methyl alcohol1 coatlaining 0-5 per celnt,. olf hydro-gen chloride and the solutions were united with Solutio'n C. Examination of Solution G.-This ext,ract tagethes with that from Residue D, was hea,ted a t looo until the robtion became const,ant after which the equilibrium mixture of the methylgluco-sides was isohted in the usual mannelr.Somelwha.t greater difficulty was experienceld in obtaining the produat in a. pure1 condit'ion and a.lthough the material beihaved shasply a,s a glucoside t,oIwards Felhling's wht,ion the melting point sholweid a wider range' than usual (105-140°). The specific ro'tahion was also slight'ly low ([a], + 111.5O in place of + 114O). Yi.edd 6.2 gra.ms (Found, OM@= 15-98 per cent.). C=43.17; H=7*16; OMe=15.22. C ~ C . C=43.30; H=7*22; Isolation of Crystalline Glucose from Cellulose. The preparations of crystlalline melt'hylglucmide olbtained it5 described were mixtures of the a- aad P-forms in equilibrium. This was confirmed by cryshllising a sample slowly from methyl alcohol. The B-form was then retrained in solution and the u-iw 1500 THE CONSTITUTION OF POLYSACCHARTDES.PART 11. meride which separated showed after drying a t 50°/25 mm., La], + 157*53 in watelr. This agrees elxactdy witlh the standard value (+ 157.6O). Moreover the material melted at 1GFi0 the melting polint being unaflelctd by admixture witlh pure1 a-methylglucosidet. Furkher proof of the stlandard of purity attltlained was aflolrded by the hydrolysis olf the mixed glumides. A 5 per cent. solution in 4 per cent(. hydrolchlotria acid was heateid at looo polarimetria reladings being tlaken ewery fifteen minutes. The permanent vdue observed when calculated for the1 weight od glucow farmed was +-53*Oo in place o t -+52*5O. After neutlralisatiom with barium arbanate and evaporation to dryness under diminished pressure it syrup mixed with barium chloride remained. This was extraded with boiliqg absolute aIcohot1 the sollutjion demlolrised f i l b r d and slowly canaent,rated at a lolw tempelratlure. On nudelatian and stirring crystalline glucose readily separated After a second crystallisation from absolute alcohol the yield olf dry sugar amoanted to 60 per cent. 09 thatl required by theory. The! glucose melted at 145O and gave glucosephenylosazone (m. p. 204-205O uncarr.) Calc. c = 40.00 ; H=6.67; OMe=O p a ct4nt.). When dissolveid in water the initial spe:cific rotation od +100*8O was relcmded and tlhis diminished to the canstlantl value1 of + 52.37O. The abwe result prolveisl conclusively that the methylglumside employed did noit contain any isomeria methylheaoside~ and that! consequeintly no mannow or gdadms residues are prewnt in celllulose. (Found C = 39.97 ; H = 6-54 ; OMe= 0. We dwim to record our tihadcs to the Royal Cbmmissionws for the 1851 Exhibition far a. Re'search SchoZa,mhip held by one of us during the progress od the above remarah. CHEMICAL RESEARCH LABORATORY, UNIVERSITY OF ST. ANDREWS. UNITED COLLEGE OF ST. SALVATOR AND ST. LEONARD, [Received November 4th 1920.