~Cellulose.~--There is little doubt that the furfural-yielding groups of the original are isolated in the form of the beta-cellulose. Tollens emphasises this fact in his studies of cellulose-estimation methods. We had previously shown (original, p. 159) that the yield of furfural is not affected by the _chlorination_, but it appears from our numbers that only 50 p.ct. of these groups remain in the isolated cellulose, the residue undergoing hydrolysis to soluble compounds. In a carefully regulated hydrolysis following the chlorination it appears that the furfuroids are almost entirely conserved in the form of a cellulose.

Moreover, an investigation of the products dissolved by sodium sulphite solution from the chlorinated fibre has shown that they are practically free from furfuroids. This enables us to exclude the furfural-yielding groups from the lignone complex. At the same time, through our later studies of the hydroxyfurfurals, it is certain that these products are represented in the fibre substance and probably in the lignone complex.

~Chlorination Statistics.~--It has been pointed out by a correspondent--to whom we express our indebtedness--that we have made a mistake in calculating the proportion of lignone from the ratio of the Cl combining with the fibre substance or lignocellulose (p.ct), to that of the Cl _present in_ the isolated lignone chloride (p.ct.). The lignocellulose combines with chlorine in the ratio 100 : 8, but the lignone chloride _containing_ 26.7 of chlorine means that, neglecting the hydrogen subst.i.tuted, 73 of lignone combine with the 27 of chlorine approximately. On the uniform percentage basis the calculated proportion of lignone would be 8/37, or a little over 20 p.ct.

In regard to the proportion of hydration attending the resolution, we have shown on const.i.tutional grounds that this must be relatively small.

a.s.suming approximately the formula C_{19}H_{22}O_{9} for the lignone residue as it exists in combination, and the anhydride formula for the cellulose, these revised statistics now appear, as regards the carbon contents of the lignocellulose:

Cellulose, 44.4 C; lignone, 57.8.

80 44.4 100 = 35.52 20 57.8 100 = 11.56 _____

47.08 p.ct. C in lignocellulose.

These conclusions are in accordance with the experimental facts, and, taken together with the new evidence we have acc.u.mulated from a study of the lignocellulose esters, we may sum up the const.i.tutional points as follows: The lignocellulose is a complex of

Cellulose alpha Cellulose beta Lignone 65 p.ct. 15 p.ct. 20 Allied to the normal Yielding furfural One-third of which celluloses approximately 50 p.ct. is of benzenoid type

The lignone contains but little hydroxyl. The celluloses are in condensed hydroxyl union with the lignone, but the combination occurs by complexes of relatively large molecular weight.

DIE CHEMIE DER LIGNOCELLULOSEN--EIN NEUER TYPUS.

W. C. HANc.o.c.k and O. W. DAHL (Berl. Ber., 1895, 1558).

~Chemistry of Lignocelluloses--A New Type.~

The stem of the aquatic _aeschynomene aspera_ offers an exceptional instance of structural modification to serve the special function of a 'float,' 1 grm. of substance occupying an apparent volume of 40-50 c.c.

This pith-like substance is morphologically a true wood (De Bary), and the author's investigations now establish that it is in all fundamental points of chemical composition a lignocellulose, although from its colour reactions it has been considered by botanists to be a cellulose tissue containing a proportion of lignified cells. Thus the main tissue is stained blue by iodine in presence of hydriodic acid (1.5 s.g.), and the colour is not changed on washing. The ordinary lignocelluloses are stained a purple brown changed to brown on washing. The reactions with phloroglucol and with aniline salts, characteristic of these compounds, is only faintly marked in the main tissue, though strongly in certain individual cells.

The following quant.i.tative determinations, however, establish the close similarity of the product to the typical lignocelluloses:

_Elementary a.n.a.lysis._--C 46.55, H 6.7. _Furfural_ 11.6 p.ct., of which there remained in the residue from alkaline hydrolysis (71 p.ct.) 8.0, i.e. about 70 p.ct. The distribution of the furfuroids is therefore not affected by the alkaline treatment.

_Chlorination._--The substance (after alkaline hydrolysis) takes up 16.9 p.ct. Cl, of which approximately one-half is converted into hydrochloric acid.

_Methoxyl._--O.CH_{3} estimated = 2.9 p.ct.

_Ferric Ferricyanide Reaction._--Increase of weight due to blue cyanide fixed (1) 75 p.ct., (2) 96 p.ct. Ratio, Fe : CN = 1 : 2, 4.

_Hydroxyl Reactions._--In the formation of nitric esters and in the sulphocarbonate reaction the substance gave results similar to those obtaining for the jute fibre.

These results establish the general ident.i.ty of this peculiar product of plant life with the lignocelluloses, at the same time that they show that certain of the colour reactions supposed to characterise the lignocelluloses are due to by-products which may or may not be present.

(p. 172) ~Composition of Elder Pith.~--In a systematic investigation of the celluloses in relation to function we shall have to give special attention to the parenchymatous tissues of all kinds. These are, for structural reasons, not easily isolated, for which reason and their generally 'inferior' functions they do not present themselves to chemical observation in the same obvious way as do their fibrous relatives. The pith of the elder, however, _is_ readily obtained in convenient ma.s.ses, and a preliminary investigation of the entire tissue has established the following points:

The _reactions_ of the tissue are in all respects those of the lignocelluloses.

_Composition._--Ash, 2.2 p.ct.; moisture in air-dry state, 12.3 p.ct.

Alkaline hydrolysis (loss): (a) 14.77, (b) 17.84. Cellulose (yield), 52.33 p.ct. Nitrate-reaction complicated by secondary reactions and yields low, 90.95 p.ct. _Sulphocarbonate reaction:_ Resists the treatment, less than 10 p.ct. pa.s.ses into solution.

_Furfural._--The original tissue yields 7.13 p.ct.; the residue from alkaline hydrolysis (b) 5.40 p.ct.

This tissue is, therefore, a lignocellulose having the chemical characteristics typical of the group, but of less resistance to hydrolytic actions.

The investigation will be prosecuted in reference to the cause of differentiation in this latter respect. Probably the pectocelluloses are represented in the tissue.

~The Insoluble Carbohydrates of Wheat (grain).~

H. C. SHERMAN (J. Amer. Chem. Soc., 1897, 291).

(p. 171) This is a study of the const.i.tuents of the cell-walls of wheat grain. Bran was taken as the most convenient form of the raw material, being freed from starch by treatment with malt extract, and further treated (1) with cold dilute ammonia, (2) cold dilute soda lye (2 p.ct.

NaOH), and (3) boiling 0.1 p.ct. NaOH. The product retained only 1.25 p.ct. proteids, and yielded 15.62 p.ct. furfural.

_Acid Hydrolysis._--The product was boiled 30 mins. with dilute acid (1.25 p.ct. H_{2}SO_{4}), and the solution boiled until the Fehling test showed no further increase of monoses. At the limit the reducing power of the dissolved carbohydrates was 91.3 p.ct., that of dextrose.

Converted into osazones the a.n.a.lysis showed them to be _pure pentosazones_. The _hemicellulose_ of wheat is, therefore, according to the author, _pure pentosane_.

_Residue._--This was a lignocellulose yielding 11.5 p.ct. furfural. It was subjected to a series of treatments with ferric ferricyanide, and the proportion of Prussian blue fixed was determined by increase of weight, viz. from 10 p.ct. to 47 p.ct. according to the conditions. The results confirmed those of Cross and Bevan first obtained with the typical lignocellulose (jute).

_Chlorination._-The residue was boiled with dilute alkali, washed, and exposed to chlorine gas. The resulting lignone chloride was isolated by solution in alcohol, &c. It yielded 26.7 p.ct. Cl on a.n.a.lysis. In this and its properties it appeared to be identical with the product isolated by Cross and Bevan from jute, with the empirical formula C_{19}H_{18}Cl_{4}O_{9}.

_Cellulose_ was isolated from the residue by three of the well-known methods, and the following comparative numbers are noteworthy:

_________________________________________________________________________

F. Schulze

Lange

Cross and

Method

Dil. HNO_{3}

Fusion KOH

Bevan

KClO_{3}

Chlorine, &c.

_____________________________

______________

____________

_______________

Cellulose p.ct. obtained

66.0

39.3-43.1

66.5

Furfural p.ct. of cellulose

7.0

3.96

5.62

Residual nitrogen

0.22

0.03

0.00

Ferricyanide reaction,

Prussian blue fixed

6.04

0.89

0.92

_____________________________

______________

____________

_______________

The author remarks: 'It is evident no one feature can be urged as a criterion in judging between the methods, but all must be taken into consideration. Such a comparison shows the superiority of the chlorination method.'

The cellulose is not of the normal (cotton) type, since on treatment with sulphuric acid it dissolves with considerable discolouration, but only to the extent of about 80 per cent. The dissolved monoses converted into osazones were found to consist of hexoses only. The cellulose treated with caustic soda solution (5 p.ct. NaOH) in the cold yielded 20 p.ct. of its weight of soluble const.i.tuents, but as the residue yielded 3.34 p.ct. furfural the attack of the alkali is by no means confined to the furfuroids.

~Animal Digestion of the Const.i.tuents of Bran.~--Observations on a steer fed upon wheat bran only established the following percentage digestion of the several const.i.tuents:

Soluble carbohydrates 96.9 Starch 100.0 Free pentosanes 60.2 Cellulose 24.8 Lignin complex 36.7 Proteid 82.96 Ether extract 42.73 _____________________ ______

Nitrogen-free extract 76.08 Crude fibre 32.21

JOURNAL OF THE IMPERIAL INSt.i.tUTE

(Research Department, Vols. 1-2, 1895-6).

(p. 109) In this journal appear a series of notices of the results of a.n.a.lyses of vegetable fibres by the method described in 'Report on Miscellaneous Fibres' (Col. Ind. Exhibition Reports, p. 368) [C. F.

Cross]. These investigations deal with the following subjects:

1895. p. 29 Various Indian Fibres--more particularly Sida.

118 (a) Fibres from Victoria; (b) Special a.n.a.lyses of (a) Samples of Jute; (c) Paper-making Fibres from S. Australia.

202 Fibres from Victoria.

287 Fibres from Victoria.

366 Sisal from Trinidad.