1410 ARMSTRONG AND HILDITCH CONVERSION OF THB 8IMPLE CXXXIT1.-Conversion of the Simple Sugars into their Enolic and Ethylene Oxide Forms. By EDIWARD FRANKLAND AWSTRONG and THOMAS PERCY HILDITCH. No group of compounds is more remarkable than that of the hexoses on account of their extreme mutability. Glucose for example may be obtained in an a- and a /3-crystalline form and if either of these is dissolved in highly purified water the solution i s all but stable. Yet if a trace of alkali is added an equilibrate BUGAM INTO THEE BNOLIC AND ETHYLENE OXIDE] FORMB. 1411 mixture of the two isodynamic forms is practioally instantaneously produced. The change has been specially studied by Lowry (T., 1903 83 1314; 1904 85 1551) and its character has been dis-cussed also by one of the authors (T.1903 83 1305). The more gradual change also under the influence of alkali, from one hexose into another within sections including as many as four members of the group (for example glucose +- - mannose - fructose glutose) was brought to light through the pains-taking labours of the late Lobry de Bruyn (Rec. trav. china. 1895, 14 156 204). In this case the transition has been assumed to involve the formation of an enolic form as an intermediate term. In the case of the hexoses the aldehydic form (aldehydrol) is commonly supposed to have but an ephemeral existence and it is held that OG and @-glucose are butylene oxide or pentaphane deriv-atives corresponding with the two methyl glucosides. The latter, i t is well known are simultaneously produced by the action of a strong acid on glucose dissolved in methyl alcohol; a residue long unexplained has recently been shown to contain a third isomeric methyl glucoside regarded by Fischer and by Irvine as the derivatives of an ethylene oxide form of the hexose.Irvine and his school (T. 1915 107 524; 1916 109 1305 etc.) have shown that the new methyl glucoside is characterised by its capacity to condense with acetone to reduce potassium perman-ganate solutions and to undergo auto-condensation and that the activity of the parent glucose from which it is derived far exceeds that of a- or &glucose. The observations we have to place on record relate to the changes effected in the simple sugars by acids and by alkalis as measured both by means of the polarimeter and by means of increased liability to oxidation.A solution of either a- or &glucose in water (both well purified) is practically unaffected .by permanganate. I n acid solution, reduction of the permanganate sets in a t once and at a definite rate under definite conditions ; thus under the standard conditions detailed in the experimental part 10 C.C. of 1 per cent. of a- or B-glucose in Rl 10-hydrochloric acid solution decolorise 2 C.C. of 3 / 100-permanganate solution a t 2 5 O in twenty-eight to thirty minutes. That the change is instantaneous is proved by the fact that the reducing power acquired is independent of the time during which the acid has aoted solutions containing acid which have been kept various times all having the ~ a m e reducing power.The strength of the acid however is a factor in the change the effect being less the weaker the acid and likewise the less concentrated the acid. * 3 Q 1412 ARMSTRONG AND HLLDITaH CONVERSION OF THE SIMPLE We coxwider that the active agent is the ethylene oxide modifi-cation of glucose, CH,(OH)*CH(OH) *CH(OH)*CH(OH)*CH*CH*OK* \/ 0 The amount of the ethylene oxide form present is regulated by an equilibrium depending on the strength of the acid but that it is small is shown by the fact that the optical rotatory power of glucose in an acid is the same as in an aqueous solution. The ethylene oxide form is reproduced as oxidation by the perman-ganate proceeds as shown by the fact that fresh additions of permanganate solution are decolorised.Mannose and fructose are affected by acids in a similar way but the altered mannose solution acts far more rapidly than that of glucose; fructose is only slightly less active than mannose. For example under precisely comparable conditions whereas the reduc-tion of permanganate by glucose occurs in twenty-eight t o thirty minutes that by mannose takes only from twelve to thirteen and that by fructose sixteen to seventeen minutes. CH,*OHt' I CH2 I\O ,C*OH CH*OH H O d 0 ] p o o/j H d \CH \ I H O ~ H H O ~ H H O ~ H H O ~ H or CH H OH H&OH H&OH ~ 1 3 - O H H &OH H C ~ O O H H ~ O H H ~ O H H ~ O H H&OH ~H,*OH ~H,*OH &H,.OH ~ T ~ . O H ~H,*OH Glucose. Mannose. L Y / Fructose. In view of the close relation of the ethylene oxide forms of these sugars it seems not improbable that one of the isodynamic forms is the more oxidisable and that this is the form present in mannose.The reduction of permanganate (as well as of methylene-blue and of indigo-blue) is also promoted by the addition of alkali. In * A t first sight the alterations involved in the conversion of butylene oxide into ethylene oxide or enolic forms of glucose appear somewhat far reaching when viewed only in the light of the conventional structural formulae. If, however structural models are prepared of the sugars on the lines of the Pope-Barlow hypothesis of close packing a more rational interpretation of the changes is realised. f This form of fructose is present in sucrose according to Haworth and Law (T. 1916 109 1314) BUUARS INTO THEIR ENOLICI AND ETHYLENE OXIDE FORMS.1413 this case however the change of the hexose is more gradual as, within limih the solution is the more active the longer the alkali has acted. The interactions in the three cases take place a t corre-sponding rates showing that in each case the same change is being studied Under the experimental conditions observed the reduc-tion phenomena correspond only with the changes in structure which take place in the course of the first few hours and it is unlikely that any far-reaching disturbance has occurred in the carbon chain beyond the atoms 1 and 2. As shown by the polarimeter the equilibrium between the a-and B-butylene oxide forms of the sugar is established instant; aneously in alkaline solution and subsequently the optical activity falls slowly but it is still of considerable magnitude a t the end of six hours.After this it continues to fall some fifteen to twenty days being required before the solution loses its activity. To judge from the slowness with which alkali acts as compared with acid taking into account the instantaneous equilibrium of the u-and &forms in the presence of alkali it is clear that the latter are not concerned in the change. It therefore seems probable so far as the effect produced by acids is concerned especially in view of the distinctly basic character of ethylene oxide that the ethylenic oxide form of the hexose not an enol is the active agent. Assuming the enol to be concerned acids equally with alkalis should convert one hexose into another in the Lobry de Bruyn change but this is known not to be the case.It is noteworthy however that ethylene oxide itself has no reducing power on either methylene-blue or indigo-blue in alkaline solution although it readily affect<s permanganate. It is by no means clear in fact that the action of alkali is comparable with that of acid and it may well be that reduction is effected by the enol. Whereas possibly in acid solution a salt of the basic ethylene oxide hexose is formed in alkaline solution the scission of all ring systems and the production of a metallic salt of the open-chain enol appears more probable. Whilst glucose decolorised the standard amount of permanganate initially in eleven minutes (in the presence of one equivalent of sodium hydroxide) fructose takes three and mannose twenty-five minutes; the figures after the solutions have remained for five hours are respectively glucose 3 fructose 1.25 and mannose 9 minutes.As the formulae show all three substances can give the same enolic form. On the supposition that “enol” rather than ‘‘ ethylene oxide ” is initially formed in the presence of alkali the configuration represented by fructose is most and that represented 3 a* 1414 BRMSTRONQ AND RUCIDITOH CIOHVIORSION OF THE EIIMPLIB by mannose least prone to undergo enolisation. contrast to the behaviour of the same three sugars towards acid. This is in marked CHO CHO CH*OH CH,*OH &OH 60 I H&OH H O ~ H H O ~ H H O ~ H H O ~ H 4- HO-OH -+ a6-OH H ~ O H H ~ O H H&OH Hh*OH H&OH H&OH ~ 6 .0 ~ I dH9*OH dH,*OH CH,mOH dH,*OH Gluooee Mannose. Common enolic Fructose. form. CH,*OH I &OH b 0 H H ~ O H Hb*OH bH,*OH Fructose alternative enol. The change in presence of alkali is qualitatively proportionate to the strength of the alkali. It is of special interest in this con-nexion that pyridine has a similar effect to the other alkalis. In this case as pyridine is such a weak alkali the sugars were dis-solved in the base itself. Decolorisation of glucose took place initially in about fifty to sixty minutes increasing to thirty minutes in about five hours whereas with fructose the times were thirty minutes initially and eight minutes after six hours. This observation is quite in harmony with the important part which pyridine and quinoline have played in sugar chemistry as the media in which epimeric changes are effected causing the re-arrangement of the groups attached to the asymmetric carbon atom at the end of the chain.Both in the interconversion of epimeric hexonio acids (E. Fischer Ber. 1890 23 2625) and of epimeric glucosides (E. Fischer and von Mechel Ber. 1916 49 2813; 1917, 60 71 l) formation of intermediate modifications must be involved. The alterations in structure which we have followed by the changee in reducing power are clearly in no way related to those known as mutarotation. Whereas the former take place instant-aneously in the presence of acid and more slowly in the presence of alkali mutarotation is brought about immediately in the presence of alkali and more slowly in the presence of acid.EXPERIMENTAL. Reducing tests were made by withdrawing 10 C.C. of the sugar solution (generally 1 per cent. or 1 / 18 gram-molecule of hexose per Iitre except when polarimetric comparisons were also made in which case 5 per cent. solutions were used) into a stoppered test-tube and adding 2 C.C. of the standard permanganate or dye solu-tion ; N / 100-potassium permanganate was invariably used whils SUQARS INTO THE33t PNOLIU AND ETRYLENE OXIDE FOBBSS. 1414 the dye solutions consisted respectively of a 0-025 per cent. solu-tion of methylene-blue and a solution of neutral indigo sulphonates which contained 0.045 per cent. of indigotin. In those cases in which it wits desired simultaneously to neutralise the acid or alkali present special solutions of N / 100-permanganate were employed containing respectively the amount of sodium hydroxide or of sulphuric acid per 2 C.C.necessary to neutralise the acid or alkali present in the 10 C.C. of sugar solution exactly. The behaviour of permanganate with the sugar solutions varies considerably according to the conditions studied ; thus acid fructose or mannose solutions pass simply from pink to clear white and perfectly definite end-points are obtainable in these cases. Ip other cases generally those of slowly reducing acid media or of rapidly reducing neutral or alkaline solutions the pink tint gives place with varying rapidity to a very pale yellow and ultimately t o the colourless condition.A definite colour standard just short of colourless was adopted here and sharp end-points could be obtained without difficulty. With slowly reducing neutral or weakly alkaline solutions however a precipitate of manganese dioxide of variable degree of fineness is apt to appear and it is difficult to determine the precise point a t which actual precipita-tion sets in; cases of this kind are denoted in the tables which follow by the addition of “ indefinite ” or “with pptn.” It is obvious that in the latter case the amount of reduction will not be so great as when the formation of the colourless manganous salt solution has occurred. I n point of fact whilst the 10 C.C. of sugar solution employed has contained 0.1 gram (1 per cent. solu-tion) or 0-5 gram (5 per cent.solution) of hexose the 2 C.C. of iV/ 100-permanganate assuming the action C,H,,O + 2 0 = C,H,(OH),*CO,H + H*CO,H to occur are capable of oxidising 0.0009 gram of hexose if the reduction proceeds to the manganous state or 0.00054 gram if the action is arrested a t the stage of manganese dioxide. In the extreme cases therefore we have measured the time of oxidation of 0.9 per cent. of the sugar present in the solution, whilst on the other hand we have dealt with the time of oxidation of as small a proportion as 0.11 per cent. or less of the sugar present. The stock solutions of sugars and also the portions undergoing tests were maintained a t 2 5 O throughout 1416 ARMSTRONQ AND HILDITOH CONVERSION OF THE SIMPLE Neutral Solutions. The dye reagents are unaffected by any of the sugars tested in neutral solution.Under the conditions described permanganate is decolorised by fructose in four or five hours whilst with either form of glucose the solution becomes orange in about six hours, and fades to a full yellow tint after twenty-four hours. The behaviour of mannose is very similar to that of glucose. These results are obtained equally with fresh solutions and those which have remained for a day a t 25O. It may be of interest to state that i f the test is conducted in N / 10-sodium chloride solution instead of in water precipitation of manganese dioxide sets in a t a much shorter time; the figures obtained for fructose a-glucose and mannose were respectively 93 90 and 115 minutes. Acid Solutions. (i) Aqueous 8oEutions.-The oxidation times are constant a t any age of the solution in the case of acids.This is illustrated by the figures for fructose u- and &glucose and mannose given in table I. The dependence of the time factor on the hydrion concentration is shown by the results in table 11 wherein only the mean figures are quoted; it may be emphasised that the agreement of the individual readings a t varying ages of the sugar solutions is in all cases as good as those given in extenso in table I. Two series of times are given for each sugar the first being for the simple acid solution; the second gives the values obtained when the test-solutions were neutralised simultaneously with the addition of permanganate. I n the case of the higher concentrations of hydrogen ions investi-gated neutralisation of the acid present effected a t the same time as the addition of permanganate does not appreciably alter the oxidation times.A t the lower concentrations of hydrogen ions and notably with the weaker acids such as phosphoric and acetic the oxidation times for the ‘‘ neutralised ” solutions are quicker than for the acid solu-tions and progressively so as the strength of the acid decreases. The identity of the oxidation times for ‘‘ neutralised ” and for acid solutions with the stronger acid solutions is quite definite and we consider that they indicate the persistence of the active form of the sugar after neutralisation of the acid has taken place; the meaning of the results with neutralised solutions of weaker acids is by no means clear although they are explicable to a certai SUQABS INTO THEIR ENQLIU AND BITHYLENE OXIDE BOBMS.1417 extent when it is borne in mind that the sodium salts of the weaker acids will be appreciably hydrolysed so that we have really passed over in this instance to a feebly alkaline solution of the sugar. The age of the solution is given in hours the first reading (0.1 hour) having been taken immediately solution was complete ; in most instances this reading was commenced about three to five minutes after the addition of acid. The times of oxidation are in minutes unless an explicit statement to the contrary is added. TABLE I. One per cent. Sugars in NI10-HC1 and N/lO-H,SO,. Fructose. a-Glucose. 15-Glucose. Mrtnnose. A A A A Age. Acid.Neutd. Acid. Neutd. Acid. Neutd. Acid. Neutd. Acid. 0.1 17 15 29 27 30 30 13 12 HCI 1-0 16 17 28 29 32 30 12 13 2.0 17 18 31 28 32 27 12 12 5.0 16 18 29 - 34 29 12 17 0.1 31 24 52 41 - - 25 21 H,SO, 1.0 30 26 61 43 - - 23 20 2.0 28 22 51 42 - - 22 20 5.0 30 26 53 48 - - 24 21 TABLB 11. One per cent. Solutions of Fructose and a-Glzccose in Acids of Varying Strength. Hydrion concen-tration. Acid. per litre. Equivs. NIlO-HCl ......... 0.0910 N/lO-H,SO ...... 0.0660 NISO-HCI ......... 0.0188 N/60-H,SO ...... 0.0100 N/10-H8P04 ...... 0.0130 N/lO-C,H,O ...... 0.0012 Fructose. - Acid. Neutd. 16 17 30 24 69 62* 76 75* 211 84* 280 20* a-Glucose. - Acid. Neutd. 29 28 52 43 118 97* 124 75* 238 150* 1,070 65* * With precipitation.The hydrion concentrations are taken from data in Landolt and Barnstein's " Physikalische-Chemische Tabellen." (ii) Alcoholic Solutions.-Some observations were made in methyl- and ethyl-alcoholic hydrogen chloride solutions in view of the considerable amount of synthetic work which has been carried out in these media. Concentrations of N/20-acid were employed in order to approximate to the 0.25 per cent. hydrogen chloride golutions which have most frequently been used by E. Fischer 1418 ABMBTBONG AND HILDITOH CONVERSION OF THE 8TMPLE Irvine and other workers with theBe reagents. Very eimilar results were obtained to thosle found in aqueous solutions. The methyl alcohol wa6 distilled over lime and then over a little fructme whilst the ethyl alcohol was twice distilled over lime and potassium permanganate.Ten C.C. of the neutral solvents caused precipitation in the permanganate test in tsix hours in the case of methyl and in sixteen hours with ethyl alcohol. N/20-Methyl-alcoholic hydrogen chloride however decolorised the permanganate in ninety-eight minutes the time with N / 20-ethyl-alcoholic hydrogen chloride being fifty-six minutes, TABLE 111. One per cent. Solutions of Fructose and a-Glucose in Alcoholic N f 20-Hydrogen C’h2ori.de. Solvent. Age. chloride ................................. 0.1 2.0 3-0 24.0 chloride ................................. 0.1 2.0 3.0 24.0 N/20-Methyl-alcoholic hydrogen N /20-E thy1 - alooholio hydrogen Fructose. 7 17 25 6 7 8 6 -a - Gluc ose.11 16 120 9 20 10 23 -Alkaline Solutions. I n alkaline solutions it is also possible t o utilise the reduction of the dyes indigo and methylene-blue; the former gives two fairly definite colour stages passing from blue through green to a clear red and then changing t o pure yellow. The times occupied from the commencement of the test in reaching the standard red and yellow tints are given in the tables under the columns headed respectively (I R ” and ‘( Y.” I n both cases the dyes are restored by contact with air and it was found desirable to fill the upper part of the test-tubes with an inert gas hydrogen being employed. It W U ~ found that the stock alkaline sugar solutions whether maintained under air or under hydrogen behaved the same towards the three oxidising agents.The result of neutralising the alkaline sugar solution when the latter has been freshly prepared is t o cause a very marked retard-Methylene-blue fades to a colourless solution ation in reducing power but after the alkaline solution has remained for ~ome hours the reducing power after aeutralisation becomes greater approaching that of the slkaline solution itself in about twenty-four hours. Thie shows that initially the change induced by alkali is reversed on neutralisation but the progressive alteration in behavioup is somewhat obscure although it may well be due to the appearance of small amounts of decomposition products resulting from more profound disturbance of the sugar molecule. It is evident however that the behaviour of neutralised alkaline solutions is quite distinct from that of neutralised acid solutions, so that it may be considered that the cause of the reducing activity of the sugar is not the same in each case.The subjoined tables give the full series of readings up to five hours for fructose a- and 15-glucose and mannose in the presence of one and of two equivalents of sodium hydroxide (table W). I n table V are recorded the initial and final valuw (five hours) for fructose and a-glucose with varying concentrations of sodium hydr-oxide and in table VI we quote similar data for these sugars in the presence of one equivalent of a number of aqueous alkalie. Correlation of the Reducing Action of the Sugars with Alterationa in Optical Rotatory Power. In order t o obtain a convenient polarimetric reading in the 2-dcm.tubes employed the behaviour of 5 per cent. solutioxuj of fructose and of a-glucose was examined in neutral acid and alkaline media. The concentrations of the acid (NIlO-HCI) and of the alkali (N/18-NaOH) and the conditions of the reducing time-testa were otherwise maintained unaltered. The alteration in concentration of the sugars involved the determiqation of the corresponding times of reaction with permanganate and the dyes, and the results of these and of the observations of optical rotatory power are collected in table VII NaOH 2.0 equiv. 1.0 NaOH equiv. 2.0 1.0 TABLE IV. One per cent. Sugar Solutions with One and Two Equivalents (a) Tests with Potassium Permngamte. Age.0-1 1.0 3.5 6.0 24.0 0.1 1.0 3.0 5.0 24.0 Age. 0.1 1.0 3.0 6.0 24.0 0 1 1.0 3.0 5.0 24-0 Fructose. 7-A&. Neutd. 1.75 65*0* 1.25 60*0* 0.75 8.0 0.60 0.5 0.76 0.2 3.00 60-0.t 2-00 -1.75 65.0 1.25 11.0 1.25 0.25 a-Glucow. 7-Alk. Neutd. 10.0 61.0* 8.0 9o.o-f 3.0 160.0t 2.0 60.0t 1-76; 1.25 11.0 180.0t 7.5 -4.5 73.0* 3.0 48-0* 2.5 2-6 (b) Tests with Indigo and iliethylene-Indigo. Indigo. Indigo. Fructose. a-Glucose. A c < 7 r- - Meth. Meth. <-'-, R. Y. Blue. R. Y. Blue. R. 1.00 2.50 1-60 4.00 6.00 7.00 -0.75 1-25 0.50 3.25 6.50 4.00 -0.50 1-00 0.40 2.50 5.00 3.00 -0.60 1.26 0.35 2.25 5.00 2.25 -1.25 3.00 2.00 2.60 5-00 3.00 -8 l4 2.25 3.50 3.00 9.0 12.5 11.0 1.50 2.75 1.25 6.5 11.5 7-0 1.26 2.50 1-00 6.0 11-5 5.0 7 1.25 2-60 1.00 5-0 11.0 4.0 6 3.25 4-75 2.00 6-0 9.0 5.5 4 * Precipitation set in.t Indefinit TABLE V. One per cent. Solutions of Fructose and a-Glucose with Varying OH ion concn. per litre. 0-0978 -0*0506 -0.0383 -0.0256 -0.0 129 -Fructose. I / 7 Indigo. - Age. R. Y. 0.1 1-00 2.50 5.0 0-60 1.26 0.1 2.25 3.50 5.0 1.25 2.60 0.1 5.0 7.5 5.0 1-25 3.0 0.1 6.0 8.5 5.0 3.0 4.5 0.1 15.0T 20.0 5.0 7.0t 10.0 * Precipitation set in. Perrnqganate Meth. - Blue. Alk. Neutd. 1-50 1.75 65-0* 0.36 0.50 0.5 3.00 3-00 60-0t 1.00 1.25 11-0 5.0 5.0 56*0* 1.0 1.5 24.0 6.0 7.0 29*0* 3.0 3.5 19.o-f 17.5 14.0 50.0* 5-0 5.0 5l*O TABLE VI. One per cent. $olutions of Fructose and a-Glucose with One OH ion concn.Alkali. parlitre NaOH 0.0606 -KOH 0.0506 -Ba(OH) 0.0478 -NH,OH 0.0011 -Fructose. Indigo. - R. Y. 2.25 3-50 1.25 2-50 4.60 6-00 0.75 2-60 5.00 6.00 1-00 1.60 83.0 t 56.0 very little action. * Precipitation set in. Meth. Blue. 3-00 1-00 5-00 0.76 5.00 1-26 150.0 98-0 88.0t 60.0 Permangmate. - Alk. Neutd. 3-00 60t 1.2s 11 4.00 120t 1.25 60 4.00 -1.26; -304t 61* 16.0 20* 42.l)t 39* 27.0 27* 70.0 19* 48.0 so b *E R In 00 M -& I 1 M 0. 1424 ARMSTRONG AND HILDITOH (3ONVERSION OF THE SIMPLE er3 rl I c? r+ u3 90 Q * 90 cn * 8 * rl c3 ? + Q Q I I i 00 0 r+ t- Age in hours. 0-1 0 5 1.0 2.0 3.0 4.0 5-0 6.0 24.0 Age in days.2 4 8 12 25 (c) Solutions in N/ 18Sodium Hydroxide. Fructose. Glucose. Indigo. Permanganate. Indigo. Permanganate. - Meth. +Y - Meth. A R. Y. Blue. A&. Neutd. [a&,. R. Y. Blue. Alk. Neutd. 2.0 3.5 3.0 3.0 21 -89.7' 12 17+ 153 154 30 - - - - 89.2 - - I I -1.75 3.5 1-5 2.5 16 88.0 9 14 12 13 26 1-75 2.5 1.5 2.0 15 86.7 - - - - -- - - - - - 6 11 7 7 16i - - - - - 3 6 4 4 15 1.25 1.75 1-25 1.5 144 84.2 - - - - -- 1.0 1.6 1.0 1.5 9Q 83.3 - - - -1-75 2.25 2.0 1.5 1-0 81.5 3 74 5 5 One per cent. Solutdons of Fructose m d Glucose in Pyridine. In view of the interest attaching to the action of pyridine on the sugars a few experimenh were made with solutions of the two sugars in pure pyridine a t 25O.The pyridine was twice distilled over lime before use and boiled a t 116-116.5°. Ten C.C. of the distilled pyridine caused no alter-ation in the blue tint of indigo or of methylene-blue (2 C.C. of the staadard dye solutions) in twenty-four hours but with 2 C.C. r f N / 100-permanganate a precipitate appeared in three hours. The 1 per cent. fructose and glucose solutions in pyridine were also without action on the dyes but results analogous to those of the weaker alkalis were obtained with permanganate. These are given in table VIII in which the age of the solutions is in hours and the dewlorisation times in minutes as usual. TABLE VIII. One per cent. Solutions of Fructose and of a-Glucose in Pyriddne. Age. Fructose. a -Glucose. 0-1 30 not quite colourless 66 with very fine precipitation.1.0 23 , Y Y 40 9Y ,P I ¶ 9 9 2.0 17 , Y9 3.5 35 Y9 1 9 9 9 Y , 8.0 30 9s 9 ) 9 ) 9Y 6-0 8 , $ 9 4 4 12 y y 7 9 Comparative Experiments with Simple Aldehydes and Deriuatives of Ethylene Oxh?e. Some experiments-were made on the reducing action of aqueous solutions of acetaldehyde n-butaldehyde acetone and epichloro-hydrin and ethylene oxide. I n neutral 1 /18th gram-molecular solution all these substances with the exception of acetone caused the pink colour of the permanganate in our standard test to dis-appear within about two hours but the resulting clear orange solu-tion underwent no further change in colour for many hours. I n the case of acetone the pink tint did not entirely vanish for several hours. The resulh of the permanganate tests in iV/10-hydrochloric acid solution are given in the next table; in the case of ethylene oxide, the exact concentrations were not accurately known but tests were made with two strengths approximately gram-molecular per litre and 1 / 18th gram-molecular per litre SUGARS INTO THEIR ENOLIO AND ETHYLENIB OXfDB FORMS.1437 TABLE IX. SoJutions of Aldehydes and of Ethylene oxides in N/ 1Q-Hydrochloric Acid. Concn. Permanganate tests. A / \ Acid. Neuldhd. 1.0 64 hrs. y y Y 7 9 Y, 3.0 2 hm. , 9 10 9 1 ?? 1.0 24 hrs. y; Y 10 Y $9 3-5 24 hm. ,) Y 10 , ,, 24-0 2& hrs. , ¶ I 10 9 1) P r Substance. htre. Age. Acetaldehyde ,.. M/18 0.1 64 hrs. Pale yellow 9 minutes with pptn. n-Butaldehyde ... M/18 0.1 34 hrs. Pale yellow 13 minutes with pptn.Epichlorohydrin M/18 0.1 140 minutes. 82 minutes. 1.5 120 s y 53 S# 4.3 90 y 35 ,Y 24.0 120 , 32 9 , 1.0 132 , 89 ,* 3.0 130 , 80 99 23.0 110 y 20 ¶ S 1.0 12 , 13 ,, 2.0 17 , 14 9 9 6.0 16 ) 14 s, 23.0 15 , 14 9, Ethylene oxide ... M/18 0.1 103 minutes. 80 minutea. M 0.1 8 minutes. 10 minutes. Table X illustrates the data obtained in N/18-aqueous sodium hydroxide ; with both aldehydes acetone and epichlorohydrin and ethylene oxide indigo passed rapidly through green to a pale yellow colour and the original blue tint could not be restored by shaking the yellow solution with air. It appears therefore that some action was proceeding in these cases beyond simple reduction to indigo-white and the colour test is not valid in this instance.TABLE X. Solutions of Aldehydes and of Ethylene Oxides in N/18&d6um Hydroxide. Concn. Permang;enate. Per Methylene-Substance. litre. Age. blue. Alkaline. Neutralised. 11* 11* 6* 60 16* 30 16* 18 6* 1 Acetaldehyde .,.,.. M/18 0.1 sot 1.0 32t 3.5 21 t 5.0 23t 23.0 351. * Precipitation set in. t Indefinite end-point 1428 CONVERSION OF THE SIMPLE SUGARS ETC. TABLE X . (continued). Solutions of Aldehydes and of Ethylene Oxides in N/ 18-Sodium Hydroxide. Concn. P r Substanoe. htre. Age. n-Butddehyde.. . . . . . . M/18 0.1 1.0 3.5 5.0 23.0 Acetone ..... ... .......... M/18 0.1 ;:;I 5.3 23.0 Epichlorohydrin . . . . . . M / 18 0.1 1.5 4.3 24.0 Ethylene oxide . . . . . . M/18 0.1 1.0 3.0 23.0 M 0.1 1.0 2.0 6.0 23.0 Methylene -blue. 20 14 12 13 .39t Permanent. Permanent. 9 9 9 9 9 9 Permanent. 9 9 9 9 9 s Permanent. 9 s 9 9 9 ) 9 9 Permanganate. Alkaline. Neutrali&d. 8* 8* 6* 8* 7* 7* 8* 8 19* 2 Pale green, then turbid yellow. 113* loo* 95* 80* 245t 236t 174t 220t 76t 7 55 19* In 16 hours, precipitation, but still pink. 207 140 110 110 352 350 340 320 87 84 88 77 63 * Precipitation set in. j- Indefinite end-point. The most interesting points in this series of experiments are: (i) The alkaline solutions of the aldehydes reduce methylene-blue similarly to the sugars and there is some similarity in their behaviour to alkaline permanganate. (ii) The ethylene oxide derivatives show no similarity to the sugars in alkaline solution either with respect t o permanganate or to methylene-blue. (iii) On the other hand the acid solutions of the ethylene oxide compounds display great likeness to those of acid sugar solutions, both in the (( acid ” and “ neutralised ” permanganate tests. WARRINUTON. [Received November 13th) 19 19.