132 POIV'VELL AND WHITTAKER THE CHEMISTRY OF LIGNIN. XXI1.-The Chemistry of Lignin. Part I I . A Comparison of Lignins Derived from Various Woods. By WALTER JAMES POWELL and HENRY WHITTAKER. IN Part I of this series (J. 1924 125 357) the preparation and properties of lignin derived from flax shoves and of several of its derivatives were described. The chief conclusions drawn were that the empirical formula for flax lignin which agrees most closely with all the analytical data is C,5H,,0,, and that the molecule contains one active aldehyde group and nine hydroxyl groups, of which four are methylated. The investigation has now been extended on the same lines to lignins obtained from a number of woods and poplar birch ash spruce larch and pine lignins, together with some derivatives of these substances have been examined.The wood lignins were isolated in the same way as flax lignin t'he wood in the form of small chips being digested for 6 to 10 hours under pressure at temperatures varying from 140" to 160" with caustic soda solution the strength of which varied from 8 to 12%. The black liquor was drained from the cellulose and treated while still hot with a slight excess of hydrochloric acid, the precipitated lignin being washed by decantation with hot dilute hydrochloric acid centrifuged and dried. An objection might be raised that tlhis method of isolation would cause internal changes to take place in the lignin molecule. Lignosulphonic acid however PART Ir. A COMPARISON OF LIGNINS ETC. 133 has been obtained from spruce wood by EorBe and Hall ( J .SOC. Chem. Ind. 1924 43 2 5 7 ~ ) by a mild process-the action of a 7% solution of sulphurous acid at 100-1 lO"-and the composition of this substance corresponds with the formula C26H30012S from which the formula C,,H,,O for lignin is obtained. This requires C = 64.4 H = 5.9% ; and these figures are in close agreement with our results (Table I). Our product is therefore not more highly condensed than that of Dorde and Hall and both are probably closely related to lignin as i t exists in the lignocellulose complex. The crude lignin was purified by pouring the solution in aqueous acetone into a large volume of hot 20% hydrochloric acid the precipitate being collected and thoroughly washed with hot water. The results now obtained show conclusively that the lignins from the different natural sources so far examined are derivatives of the same hydroxy-compound and that they differ only in the number of methoxyl groups which they contain.This difference is mainly due to the fact that the methoxyl content of lignin in wood varies from one species to another while in any one species variations occur according to the age of the tree. I n addition the methoxyl content of lignin isolated in the manner already described is affected by the strength of caustic soda solution and the time and tem-perature of digestion. Lignin as isolated is therefore a mixture of methylated derivatives of a polyhydroxy-compound and i t should be possible by chemical means to effect separation into iractions of definite methoxyl content but our attempts to do so have not yet met with success.The separate fractions obtained by precipitating lignin from its solution in caustic soda by addition of hydrochloric acid gave the same figures on analysis. I n deducing the formula c4,H480,6 from the analytical results given by flax lignin and its derivatives the assumption was made that the molecule contained a definite number of methoxyl groups, vhereas in our present view there is no reason to suppose that such is the case. Instead of the total of nine hydroxyl groups indicated 11y analysis of acetyl-lignin i t is possible to suggest forrnulx! con-taining eight or ten hydroxyl groups (e.g. C,,H,,O1,) which agree equally well with the analytical data. At this stage of the work, however i t is not desirable to make a decision in favour of one or the other of these formula and until more data are available we intend to continue using the formula adopted for flax lignin, From the six wood lignins purified as described above the acetyl compounds were prepared by treatment on the steam-bath with acetic anhydride and sulphuric acid.The acetyl content varied according to the methoxyl content but in each case using the '4 5"48O 16 134 POWELL AND WHITTAKER THE CHEMISTRY OF LIGNIN. empirical formula C45H48016 as the basis of calculation the total of acetylated and methylated hydroxyl groups was nine although the actual number of methylated groups varied between three and five. All the acetyl compounds were insoluble in cold alkali and i t is therefore unlikely that there are any carboxyl groups in the molecule.The variations in methoxyl content of the various wood lignins prepared were not sufficient to affect the determinations of carbon and hydrogen which were carried out on the purified samples and the figures obtained were constant within the limits of experimental error. Table I gives a summary of the analytical results and also the number of groups calculated on the basis of the formula C4,H,,0, for the parent hydroxyl compound itself. It is proposed to assign the name ZignoZ to this compound. Substance Determin-analysed. ation. Lignin. C H CHO Acetyl-lignin. OMe COCH, No. of OMe groups. No. of COaCH groups. Total groups. Acetylmethyl- OMe No. of OMe groups. No. of COCH groups.Total groups. lignin COCH, TABLE I. Source of Lignin. r h -. Flax. Larch. Pine. Spruce. Ash. Birch. Poplar. 63.9 63.8 63.4 64.0 63.2 63.2 63.3 5.8 5.2 5.6 5.5 5-6 5.5 5-8 11.8 9.0 11-5 11.0 13.3 15.2 12.6 20.6 23.0 18.9 19.4 17-6 14.5 17.5 4.0 3.1 3.9 3-8 4.5 5.0 4.3 5.0 6-8 4.5 4.8 4.3 3.4 4.3 9.0 8.9 8.4 8.6 8.8 8.4 8.6 20-3 19.9 20.4 23.1 20.0 22.4 22.7 11.1 11.8 10.5 8.0 11.3 8.1 7.5 6.5 6.4 6.5 7.2 6.4 7.0 7.0 2.6 2.7 2.4 1.8 2.6 1.8 1.7 9.1 9-1 8.9 9.0 9.0 8.8 8.7 3.1 - 2.9 3.1 - 3.2 -The table shows that lignol from each of the wood lignins examined has the same empirical formula and the same number of hydroxyl groups as that from flax lignin. In order to verify this statement experimentally attempts have been made to isolate lignol from the various lignins by the action of hydriodic acid.However on treating lignin with hydriodic acid (d 1.7) at 130° a dark brown product which appeared to be a mixture was obtained. The main fraction of this product wm insoluble in caustic soda and cannot therefore be lignol and as the percentage of carbon found is much higher than that calculated i t appears that reduction takes place during the reaction with hydriodic acid. Confirmation of the results and conclusions drawn from the analysis of the acetylated lignins was obtained by examination of the methylated derivatives. The latter were prepared by treatment of the sodium hydroxide solution of lignin with methyl sulphate in the cold. It is not possible to methylate more than seven of the hydroxyl groups under the conditions employed an PART 11.A COMPARISOX OF LIGNINS ETC. 133 in some cases full methylation was obtained only after several treatments. The product from a single treatment however after being washed with dilute sodium hydroxide solution was acetylated in the usual manner and the total of methoxyl and acetyl groups in all the products was found t o be practically constant. It was stated in the paper on flax lignin (Zoc. cit.) that the proportion of acetyl in acetj-lmethyl-lignin was too low to account for the presence of two acetyl groups. In the experiments on vhich this statement was based acetic anhydride alone was used as the acetylating agent, but i t has since been found that in presence of sulphuric acid the diacetyl derivative is readily formed.The analytical data given in Table I for the acetylmethyl-lignins strongly support the view that all the various lignins examined have the same empirical composition and number of hydroxyl groups in the molecule. Derivatives of an entirely different character from those already described result when lignin is treated with chlorine or bromine, and if there were any fundamental differences in the molecular structure of the various lignins i t is unlikely that the same halogen-ated derivative would be obtained in each case. In our study of flax lignin we found that the chlorination or bromination product contained twelve halogen atoms and that a large quantity of hydrogen chloride or bromide was evolved during halogenation. Most of the methoxyl groups originally present in the lignin were removed chlorolignin apparently containing two and bromolignin containing only one methoxyl group.Six of the twelve chlorine atoms in flax chlorolignin were eliminated with great ease for example by clissolving the fully chlorinated derivative in cold dilute sodium hydroxide the hexachlorolignin being precipitated on acidification. Hexachlorolignin contains only one methoxyl group and therefore the fully chlorinated product dodecachloro-lignin can contain only one such group and is similar to the brorno-compound in that respect. The higher figure €or methoxyl obtained on analysis is due t o the ease with which hydrogen chloride is eliminated during estimation in the Zeisel apparatus. The Perkiii method for the determination of acetyl also gives high results in the case of acetyldodecachloroligi~in owing to the liberation of ethyl chloride.The six u-ood lignins have been brominated and subsequently acetylated . The dodecabromolignins and their acetyl derivatives 011 analysis gave results agreeing closely with those obtained from the corresponding products derived from flax lignin (Table 11). The dodecachlorolignins prepared from two wood lignins were similar in composition to flax dodecachlorolignin. A mixturc o 136 POWELL AND WHITTAKER THE CHEMISTRY OF LIGNIN. the six wood dodecachlorolignins on solution in dilute sodium hydroxide solution and reprecipitation with hydrochloric acid, gave a product comparable in all respects with flax hexachlorolignin.It may also be noted that the various acetylated dodecachlorolignins were more readily hydrolysed than the corresponding bromo-compounds a behaviour similar to that of the derivatives from flax lignin and that in all cases both acetyl derivatives were insoluble in cold sodium hydroxide solution. A third type of derivative nitrolignin has been prepared from each wood lignin and the products were identical with that from flax lignin in composition three nitro-groups having been introduced into the molecule. As in the case of flax nitrolignin a part of the nitrogen is present as organic nitrate as is evident by the fact that a proportion can be estimated by the Lunge or Schultze-Tremann method. The total nitrogen however can only be determined by the Dumas or Kjeldahl method.The nitration of lignin proceeds equally readily whether nitric acid alone (40%) or a mixture of nitric and sulphuric acids is used the products containing the same percentage of nitrogen. These facts indicate the prob-ability of the presence of aromatic nuclei in lignin. Acetylnitrolignin is soluble in cold alkalis and it is possible that nitrolignin conta,ins one or more carboxyl groups. In this connexion it may be noted that elementary analysis of the nitro-compounds shows that some oxidation has taken place in addition to nitration. Owing to the ready hydrolysis of the acetylnitrolignins it is difficult to deduce the number of hydroxyl groups in the molecule from analyses of this compound. TABLE 11. Substance analysed. Dodecabromo-lignin Acetyldodeca-bromolignin Dodecachloro-lignin Ni trolignin Hexachloro -lignin Methyl-lignin-phenylhydrazone N - 7.5* * Samples prepared from a mixture of equal quantities of the wood lignins.Source of Lignin. Detemin- -. -~ -..- ~ _ _ ation. Flax. Larch. Pine. Spruce. Ash Birch. Poplar. 28.0 28.2 - - 28.1 - -1.5 1.6 - - 1.6 - -55.2 55.0 54.8 55.2 55.1 54.7 54-9 1.8 1.6 - - - 2.4 2.1 /OMe \CO-CH 9.4 9.8 9.0 9.8 - - 9.2 {E c1 35.1 35.5 - - - 36.0 -{OMe 5.2 4.9 - - - 5.1 -50.4 50.2 - 50.6 3.8 3.8 - 3.9 4.2 4.1 4.1 - 4.3 - 4.2 3.0 3.2 2.6 3.1 3.0 3.3 -(: 50.1 49.5* 3.0 3*2* 3-45 3*5* 20.8 204* 1 :M PART 11. A COMPARISON OF LIQNINS ETC. 137 Some discussion has recently taken place in the literature as to whether or not lignin evolves furfuraldehyde on distillatmion with 12% hydrochloric acid and in view of this we have carefully examined our products using the volumetric method for the estim-ation of pentosans in wood cellulose previously described by us ( J .SOC. Chem. Ind. 1924 43 35T). Average results for the six crude lignins obtained by acidification of the alkaline liquor showed the presence of 0.9% of pentosans. This value was reduced to 0.3% by one purification from acetone and hydrochloric acid whilst, the product from a second purification gave no trace of furfur-aldehyde. We therefore regard as untenable the view of Hagglund (Cellulosechemie 1923 4 73) that lignin contains 5% of furfural-yielding carbohydrate as an integral part of the molecule. Of the sixteen oxygen atoms present in the molecule of lignin, nine are in the form of hydroxyl groups and one as an aldehyde group.The presence of the latter was indicated by measurements of the amount of Fehling’s solution reduced by lignin and was confirmed in the case of flax lignin by a volumetric determination of the amount of phenylhydrazine required to form the phenyl-hydrazone. The following experiment shows however that the monohydrazone is only formed under the mild conditions used in the volumetric method and that the rgaction can be made to proceed further. Methyl-lignin was allowed to react with a hot alcoholic solution of phenylhydrazine and the product freed from phenylhydrazine by several precipitations from acetone and hydro-chloric acid. Analysis showed that three phenylhydrazine molecules had reacted with each molecule of lignin and therefore that two ketonic groups must also be present.Methyl-lignin was used instead of lignin in order to avoid the formation of a phenyl-hydrazine ester. These results may be summarised by writing the formula for ljgnol in the extended form C,,H,,O,(CO),(CHO)(OH),. The details of. the experimental work are similar to those given in Part I for flax lignin and are therefore not repeated. The results are published by permission of the Director of Artillery, RESEARCH DEPARTMENT, to whom our thanks are due. ROYAL ARSENAL WOOLWICN. [Received October 21st 1924. 132 POIV'VELL AND WHITTAKER THE CHEMISTRY OF LIGNIN. XXI1.-The Chemistry of Lignin. Part I I . A Comparison of Lignins Derived from Various Woods.By WALTER JAMES POWELL and HENRY WHITTAKER. IN Part I of this series (J. 1924 125 357) the preparation and properties of lignin derived from flax shoves and of several of its derivatives were described. The chief conclusions drawn were that the empirical formula for flax lignin which agrees most closely with all the analytical data is C,5H,,0,, and that the molecule contains one active aldehyde group and nine hydroxyl groups, of which four are methylated. The investigation has now been extended on the same lines to lignins obtained from a number of woods and poplar birch ash spruce larch and pine lignins, together with some derivatives of these substances have been examined. The wood lignins were isolated in the same way as flax lignin t'he wood in the form of small chips being digested for 6 to 10 hours under pressure at temperatures varying from 140" to 160" with caustic soda solution the strength of which varied from 8 to 12%.The black liquor was drained from the cellulose and treated while still hot with a slight excess of hydrochloric acid, the precipitated lignin being washed by decantation with hot dilute hydrochloric acid centrifuged and dried. An objection might be raised that tlhis method of isolation would cause internal changes to take place in the lignin molecule. Lignosulphonic acid however PART Ir. A COMPARISON OF LIGNINS ETC. 133 has been obtained from spruce wood by EorBe and Hall ( J . SOC. Chem. Ind. 1924 43 2 5 7 ~ ) by a mild process-the action of a 7% solution of sulphurous acid at 100-1 lO"-and the composition of this substance corresponds with the formula C26H30012S from which the formula C,,H,,O for lignin is obtained.This requires C = 64.4 H = 5.9% ; and these figures are in close agreement with our results (Table I). Our product is therefore not more highly condensed than that of Dorde and Hall and both are probably closely related to lignin as i t exists in the lignocellulose complex. The crude lignin was purified by pouring the solution in aqueous acetone into a large volume of hot 20% hydrochloric acid the precipitate being collected and thoroughly washed with hot water. The results now obtained show conclusively that the lignins from the different natural sources so far examined are derivatives of the same hydroxy-compound and that they differ only in the number of methoxyl groups which they contain.This difference is mainly due to the fact that the methoxyl content of lignin in wood varies from one species to another while in any one species variations occur according to the age of the tree. I n addition the methoxyl content of lignin isolated in the manner already described is affected by the strength of caustic soda solution and the time and tem-perature of digestion. Lignin as isolated is therefore a mixture of methylated derivatives of a polyhydroxy-compound and i t should be possible by chemical means to effect separation into iractions of definite methoxyl content but our attempts to do so have not yet met with success. The separate fractions obtained by precipitating lignin from its solution in caustic soda by addition of hydrochloric acid gave the same figures on analysis.I n deducing the formula c4,H480,6 from the analytical results given by flax lignin and its derivatives the assumption was made that the molecule contained a definite number of methoxyl groups, vhereas in our present view there is no reason to suppose that such is the case. Instead of the total of nine hydroxyl groups indicated 11y analysis of acetyl-lignin i t is possible to suggest forrnulx! con-taining eight or ten hydroxyl groups (e.g. C,,H,,O1,) which agree equally well with the analytical data. At this stage of the work, however i t is not desirable to make a decision in favour of one or the other of these formula and until more data are available we intend to continue using the formula adopted for flax lignin, From the six wood lignins purified as described above the acetyl compounds were prepared by treatment on the steam-bath with acetic anhydride and sulphuric acid.The acetyl content varied according to the methoxyl content but in each case using the '4 5"48O 16 134 POWELL AND WHITTAKER THE CHEMISTRY OF LIGNIN. empirical formula C45H48016 as the basis of calculation the total of acetylated and methylated hydroxyl groups was nine although the actual number of methylated groups varied between three and five. All the acetyl compounds were insoluble in cold alkali and i t is therefore unlikely that there are any carboxyl groups in the molecule. The variations in methoxyl content of the various wood lignins prepared were not sufficient to affect the determinations of carbon and hydrogen which were carried out on the purified samples and the figures obtained were constant within the limits of experimental error.Table I gives a summary of the analytical results and also the number of groups calculated on the basis of the formula C4,H,,0, for the parent hydroxyl compound itself. It is proposed to assign the name ZignoZ to this compound. Substance Determin-analysed. ation. Lignin. C H CHO Acetyl-lignin. OMe COCH, No. of OMe groups. No. of COaCH groups. Total groups. Acetylmethyl- OMe No. of OMe groups. No. of COCH groups. Total groups. lignin COCH, TABLE I. Source of Lignin. r h -. Flax. Larch. Pine.Spruce. Ash. Birch. Poplar. 63.9 63.8 63.4 64.0 63.2 63.2 63.3 5.8 5.2 5.6 5.5 5-6 5.5 5-8 11.8 9.0 11-5 11.0 13.3 15.2 12.6 20.6 23.0 18.9 19.4 17-6 14.5 17.5 4.0 3.1 3.9 3-8 4.5 5.0 4.3 5.0 6-8 4.5 4.8 4.3 3.4 4.3 9.0 8.9 8.4 8.6 8.8 8.4 8.6 20-3 19.9 20.4 23.1 20.0 22.4 22.7 11.1 11.8 10.5 8.0 11.3 8.1 7.5 6.5 6.4 6.5 7.2 6.4 7.0 7.0 2.6 2.7 2.4 1.8 2.6 1.8 1.7 9.1 9-1 8.9 9.0 9.0 8.8 8.7 3.1 - 2.9 3.1 - 3.2 -The table shows that lignol from each of the wood lignins examined has the same empirical formula and the same number of hydroxyl groups as that from flax lignin. In order to verify this statement experimentally attempts have been made to isolate lignol from the various lignins by the action of hydriodic acid. However on treating lignin with hydriodic acid (d 1.7) at 130° a dark brown product which appeared to be a mixture was obtained.The main fraction of this product wm insoluble in caustic soda and cannot therefore be lignol and as the percentage of carbon found is much higher than that calculated i t appears that reduction takes place during the reaction with hydriodic acid. Confirmation of the results and conclusions drawn from the analysis of the acetylated lignins was obtained by examination of the methylated derivatives. The latter were prepared by treatment of the sodium hydroxide solution of lignin with methyl sulphate in the cold. It is not possible to methylate more than seven of the hydroxyl groups under the conditions employed an PART 11. A COMPARISOX OF LIGNINS ETC. 133 in some cases full methylation was obtained only after several treatments.The product from a single treatment however after being washed with dilute sodium hydroxide solution was acetylated in the usual manner and the total of methoxyl and acetyl groups in all the products was found t o be practically constant. It was stated in the paper on flax lignin (Zoc. cit.) that the proportion of acetyl in acetj-lmethyl-lignin was too low to account for the presence of two acetyl groups. In the experiments on vhich this statement was based acetic anhydride alone was used as the acetylating agent, but i t has since been found that in presence of sulphuric acid the diacetyl derivative is readily formed. The analytical data given in Table I for the acetylmethyl-lignins strongly support the view that all the various lignins examined have the same empirical composition and number of hydroxyl groups in the molecule.Derivatives of an entirely different character from those already described result when lignin is treated with chlorine or bromine, and if there were any fundamental differences in the molecular structure of the various lignins i t is unlikely that the same halogen-ated derivative would be obtained in each case. In our study of flax lignin we found that the chlorination or bromination product contained twelve halogen atoms and that a large quantity of hydrogen chloride or bromide was evolved during halogenation. Most of the methoxyl groups originally present in the lignin were removed chlorolignin apparently containing two and bromolignin containing only one methoxyl group.Six of the twelve chlorine atoms in flax chlorolignin were eliminated with great ease for example by clissolving the fully chlorinated derivative in cold dilute sodium hydroxide the hexachlorolignin being precipitated on acidification. Hexachlorolignin contains only one methoxyl group and therefore the fully chlorinated product dodecachloro-lignin can contain only one such group and is similar to the brorno-compound in that respect. The higher figure €or methoxyl obtained on analysis is due t o the ease with which hydrogen chloride is eliminated during estimation in the Zeisel apparatus. The Perkiii method for the determination of acetyl also gives high results in the case of acetyldodecachloroligi~in owing to the liberation of ethyl chloride.The six u-ood lignins have been brominated and subsequently acetylated . The dodecabromolignins and their acetyl derivatives 011 analysis gave results agreeing closely with those obtained from the corresponding products derived from flax lignin (Table 11). The dodecachlorolignins prepared from two wood lignins were similar in composition to flax dodecachlorolignin. A mixturc o 136 POWELL AND WHITTAKER THE CHEMISTRY OF LIGNIN. the six wood dodecachlorolignins on solution in dilute sodium hydroxide solution and reprecipitation with hydrochloric acid, gave a product comparable in all respects with flax hexachlorolignin. It may also be noted that the various acetylated dodecachlorolignins were more readily hydrolysed than the corresponding bromo-compounds a behaviour similar to that of the derivatives from flax lignin and that in all cases both acetyl derivatives were insoluble in cold sodium hydroxide solution.A third type of derivative nitrolignin has been prepared from each wood lignin and the products were identical with that from flax lignin in composition three nitro-groups having been introduced into the molecule. As in the case of flax nitrolignin a part of the nitrogen is present as organic nitrate as is evident by the fact that a proportion can be estimated by the Lunge or Schultze-Tremann method. The total nitrogen however can only be determined by the Dumas or Kjeldahl method. The nitration of lignin proceeds equally readily whether nitric acid alone (40%) or a mixture of nitric and sulphuric acids is used the products containing the same percentage of nitrogen.These facts indicate the prob-ability of the presence of aromatic nuclei in lignin. Acetylnitrolignin is soluble in cold alkalis and it is possible that nitrolignin conta,ins one or more carboxyl groups. In this connexion it may be noted that elementary analysis of the nitro-compounds shows that some oxidation has taken place in addition to nitration. Owing to the ready hydrolysis of the acetylnitrolignins it is difficult to deduce the number of hydroxyl groups in the molecule from analyses of this compound. TABLE 11. Substance analysed. Dodecabromo-lignin Acetyldodeca-bromolignin Dodecachloro-lignin Ni trolignin Hexachloro -lignin Methyl-lignin-phenylhydrazone N - 7.5* * Samples prepared from a mixture of equal quantities of the wood lignins.Source of Lignin. Detemin- -. -~ -..- ~ _ _ ation. Flax. Larch. Pine. Spruce. Ash Birch. Poplar. 28.0 28.2 - - 28.1 - -1.5 1.6 - - 1.6 - -55.2 55.0 54.8 55.2 55.1 54.7 54-9 1.8 1.6 - - - 2.4 2.1 /OMe \CO-CH 9.4 9.8 9.0 9.8 - - 9.2 {E c1 35.1 35.5 - - - 36.0 -{OMe 5.2 4.9 - - - 5.1 -50.4 50.2 - 50.6 3.8 3.8 - 3.9 4.2 4.1 4.1 - 4.3 - 4.2 3.0 3.2 2.6 3.1 3.0 3.3 -(: 50.1 49.5* 3.0 3*2* 3-45 3*5* 20.8 204* 1 :M PART 11. A COMPARISON OF LIQNINS ETC. 137 Some discussion has recently taken place in the literature as to whether or not lignin evolves furfuraldehyde on distillatmion with 12% hydrochloric acid and in view of this we have carefully examined our products using the volumetric method for the estim-ation of pentosans in wood cellulose previously described by us ( J .SOC. Chem. Ind. 1924 43 35T). Average results for the six crude lignins obtained by acidification of the alkaline liquor showed the presence of 0.9% of pentosans. This value was reduced to 0.3% by one purification from acetone and hydrochloric acid whilst, the product from a second purification gave no trace of furfur-aldehyde. We therefore regard as untenable the view of Hagglund (Cellulosechemie 1923 4 73) that lignin contains 5% of furfural-yielding carbohydrate as an integral part of the molecule. Of the sixteen oxygen atoms present in the molecule of lignin, nine are in the form of hydroxyl groups and one as an aldehyde group. The presence of the latter was indicated by measurements of the amount of Fehling’s solution reduced by lignin and was confirmed in the case of flax lignin by a volumetric determination of the amount of phenylhydrazine required to form the phenyl-hydrazone. The following experiment shows however that the monohydrazone is only formed under the mild conditions used in the volumetric method and that the rgaction can be made to proceed further. Methyl-lignin was allowed to react with a hot alcoholic solution of phenylhydrazine and the product freed from phenylhydrazine by several precipitations from acetone and hydro-chloric acid. Analysis showed that three phenylhydrazine molecules had reacted with each molecule of lignin and therefore that two ketonic groups must also be present. Methyl-lignin was used instead of lignin in order to avoid the formation of a phenyl-hydrazine ester. These results may be summarised by writing the formula for ljgnol in the extended form C,,H,,O,(CO),(CHO)(OH),. The details of. the experimental work are similar to those given in Part I for flax lignin and are therefore not repeated. The results are published by permission of the Director of Artillery, RESEARCH DEPARTMENT, to whom our thanks are due. ROYAL ARSENAL WOOLWICN. [Received October 21st 1924.