About 0.5 gram was boiled with a small quantity of water, cooled, and a few drops of strong aqueous caustic potash or soda added, any rise of temperature being avoided; the solution was made up to 25 c.c., filtered, and examined in a 100 mm. tube. The mean of six determinations from different preparations gave the value [a] −44·1°.

Hydrolysis of Lævulomannan.

About 0.5 gram of the substance was dissolved under the conditions mentioned above, 2 c.c. of concentrated hydrochloric acid were added, and the solution heated in a bath of boiling water for 15 minutes, cooled, made up to 25 c.c., filtered and examined at 20° in a 200 mm. tube. The mean specific rotatory power of the mixed sugars after inversion was [a] +9.5° at 20°, the limits being within 1° of this value. Taking the value [a] +14·25° for mannose, and [a] -88.7° for lævulose,* a mixture of 20 parts of mannose with 1 part of lævulose would have the rotatory power [a] +9.35°, which closely approximates to the experimental value. We do not, however, wish to insist that on acid hydrolysis, lævulomannan yields a mixture of lævulose and mannose in the proportions just mentioned, as the lævulose is present in far too small a quantity to allow of accurate measurement.

Several estimations of mannose were made in the products of hydrolysis of laevulomannan by treating with phenylhydrazine acetate and weighing the phenylhydrazone formed. The following are the results obtained:

Inverted lævulomannan. Mannose phenylhydrazone.

Mannose per cent.

87.9

88.7

89.3

91.1

The determinations were carried out with the object of estimating the mannose only approximately, and as no correction was made for the solubility of the phenylhydrazone in the mother liquors, the results are probably low; they serve, however, to establish the fact that lævulomannan yields at least 90 per cent. of mannose on hydrolysis as against 95 per cent. indicated by the rotatory power of the inverted solution.

Dibenzoyllævulomannan, CH,О5(CO C6H5)2, prepared by acting on

According to Hönig, Schubert, and Jesser (Wien. Akad. Ber.. 97, ii, 534 ; Ber. Ref., 1887, 20, 721), the specific rotatory power of lævulose at 20° is expressed by the formula: [a]=-113 9635° +0.2583 q, where q represents the percentage of water in the solution. In solutions of the inverted lævulomannan of the concentration employed by us, [a], would therefore be about -88.7°.

lævulomannan in alkaline solution with benzoyl chloride, was purified by solution in glacial acetic acid and precipitation with water, this treatment being thrice repeated. The substance is a white powder quite devoid of crystalline structure; it softens at 165-170°, and is soluble in benzene, alcohol, or glacial acetic acid. On analysis: 0.2349 gave 0.5803 CO, and 0·1048 H2O. C=64·9; H=4·70. C20H1807 requires C=64.8. H=4.81 per cent.

Two determinations of the specific rotatory power in glacial acetic acid solution gave [a] - 76·0° as a mean value.

In conclusion, we wish to express our thanks to Mr. J. F. Briggs for having kindly supplied us with a quantity of the Indian clearing

nut.

LABORATORY OF THE BEETROOT SUGAR ASSOCIATION.

LIX.-The Dissociation Constant of Azoimide (Hydrazoic Acid).

By CHARLES ALFRED WEST, Assoc. R.C.S.

In view of the remarkable properties of hydrazoic acid, it was thought that a determination of the avidity of the acid would be of interest. While the work was in progress, however, it was found that Ostwald had already determined the electrical conductivity of the acid and had ascertained that it was a somewhat stronger acid than acetic acid (compare Curtius, J. pr. Chem., 1891, [ii], 43, 207). Professor Ostwald has been good enough to inform the author that the details of his work were never published and that they have not been preserved. In these circumstances, an account of the results obtained in the present experiments may be put on record.

Preparation of the Acid and Sodium Salt.

The sodium salt of hydrazoic acid was prepared by the method described by W. Wislicenus (Ber., 1892, 25, 2084). The product was dissolved in some of the water which had been prepared for the conductivity experiments, sulphuric acid added in slight excess, and the hydrazoic acid distilled off. Practically the whole had passed over when about one-third of the liquid had been distilled. The acid was estimated by titration with W/10 caustic soda, using phenolphthalein as indicator.

About 30 c.c. of the N/2 acid were neutralised with pure sodium hydroxide, then a little more of the acid was added, and the solution boiled for some time to expel the excess. The resulting solution was 3 с

VOL. LXXVII.

not alkaline to phenolphthalein, but a drop placed on neutral litmus paper gradually, but very slowly, turned the litmus distinctly blue.

Estimation of the Sodium Salt.

As hydrazoic acid distils very readily, it was thought that the sodium salt might be estimated by adding excess of N/10 sulphuric acid, boiling off the hydrazoic acid and titrating the excess of sulphuric acid with soda.

This method of estimation was tested and found to be satisfactory by titrating a solution of hydrazoic acid with N/10 soda, adding excess of N/10 sulphuric acid, boiling for 20 minutes, and then titrating back with soda. The volumes of the soda and sulphuric acid solutions used were exactly equal, showing that all the hydrazoic acid had been expelled by boiling.

Conductivity Experiments.

The water used was prepared in a well ventilated laboratory, where no other work was going on, by distilling the laboratory distilled water first with sulphuric acid, then with baryta, and then alone. The water thus prepared had a conductivity of 5.85 × 10-6 Siemens' units. This was taken into account in the measurements below.

The conductivities were determined by Kohlrausch's method, by means of induction currents and a telephone. A covered cell with platinised platinum electrodes was used, the cell being immersed to within th of an inch of the top in water at 25°.

To test the working of the apparatus, the conductivity of acetic acid was determined with the following results :

The conductivity of sodium hydrazoate was then determined with the following results:

Moo will therefore have a value of about 109. The rate of increase in the conductivity of the N/100 solution of the sodium salt with increase in temperature was determined between 15° and 35°, and was found to be 2.7 per cent. per degree.

Conductivity of Hydrazoic Acid.

Taking μ=109 for the sodium salt of hydrazoic acid, subtracting from this 44.5 for the velocity of the sodium ion, and adding 320.5 for hydrogen, we get 385 as the value of μo for hydrazoic acid.

Measurements for the conductivity of hydrazoic acid were made with the following results:

The value of k was also deduced from the rate of inversion of canesugar induced by the acid, and was found to be 0.0000186.

The dissociation of acetic acid in N/10 solution at 25° is 0·0133, and the value of k deduced from Ostwald's measurements is 0.0000180. The dissociation of hydrochloric acid in N/10 solution at 18° is 0.914. The avidity of hydrazoic acid is therefore a little greater than that of acetic acid, and only about one-seventieth of that of hydrochloric acid.

These determinations were carried out at the suggestion of Prof. Tilden, to whom the author is indebted for advice during the progress of the work.

ROYAL COLLEGE OF SCIENCE, LONDON,

SOUTH KENSINGTON, S. W.

LX.-New Glucoside from Willow Bark.

By HOOPER ALBERT DICKINSON JOWETT, D.Sc.

Ir has been generally assumed that the different species of willow bark contain the same glucoside-salicin—although in varying amount. Beck, for example (Amer. J. Pharm., 1891, 63, 581), gives the results of the assay of different species of willow bark, but no proof that the same glucoside was obtained from all the varieties. Having had occasion to examine a bark purchased as black willow, I found that the crystalline principle obtained from it by the usual method for preparing salicin, was not that substance, but a new glucoside for

which the name of salinigrin is provisionally proposed. Unfortunately, it has not been possible to determine with certainty from what species of Salix the bark was obtained, for the identification of the different species of willow, even in living plants, is admitted by botanists to be a most difficult task. Beck (loc. cit.) gives the yield of salicin from Salix nigra as 0.73 per cent. The determination of the constitution of this new glucoside has revealed an extremely interesting fact, for whilst salicin is the glucoside of o-hydroxybenzyl alcohol, salinigrin is the glucoside of m-hydroxybenzaldehyde, a substance not met with hitherto in plants.

A proximate analysis of the bark showed that, in addition to salinigrin, it contains a large amount of tannin, colouring matter, and the ordinary plant constituents. The amount of extractive matter obtained by means of alcohol and of water was 8.5 and 9.5 per cent. respectively.

Preparation of Salinigrin.

The coarsely powdered bark was boiled with a convenient quantity of water for 2 hours, the decoction strained, and the marc pressed. The aqueous extract was then evaporated to a somewhat low bulk, 10 per cent. of lead acetate added, the mixture boiled for a few minutes, and set aside. It was then strained through calico, the excess of lead acetate removed by hydrogen sulphide, and the mixture filtered. It was finally neutralised with ammonia, evaporated to a low bulk, and set aside to crystallise. The crystals were then drained by the aid of the pump, recrystallised from hot water until white, and finally recrystallised from hot alcohol. The yield of purified glucoside was about 1 per cent. of the bark, or somewhat higher than the figure given by Beck for salicin.

Properties of Salinigrin.

The glucoside separates in needles or rosettes, which, however, coalesce during collection and draining; if crystallised from water, it forms a crystalline mass; better crystals can be obtained from alcohol. The purity of the product was determined by the fact that, after five crystallisations from water, once from acetone and from alcohol, the melting point was unchanged. The crystals are anhydrous and melt at 195° (corr.) without decomposition, solidifying to a crystalline mass on cooling. It is fairly soluble in cold, and exceedingly soluble in hot water, sparingly soluble in cold, but more so in hot alcohol, also sparingly soluble in hot acetone, and almost insoluble in ether, light petroleum, or chloroform. Determinations of its solubility in water and in alcohol gave the following results: