in water, but only sparingly so in dilute alcohol. It is not sensitive to light, and can be heated to 100° without decomposition.

Attempts to prepare a corresponding compound of tin were unsuccessful.

We are indebted to Mr. W. S. Denham for valuable assistance in part of this work..

CHEMICAL LABORATORY,

GLASGOW AND WEST OF SCOTLAND TECHNICAL COLLEGE.

LV.-Derivatives of aa'-Dibromocamphorsulphonic Acid. By ARTHUR LAPWORTH.

THE substitution derivatives of camphor which have hitherto been obtained are of at least three distinct types. In the commonest type, the representatives of which are known as a-derivatives, the substituent replaces a hydrogen atom of the -CH, CO- group, and both mono- and di-derivatives of this class are known; they are produced by the action of chlorine and bromine at fairly low temperatures, by nitration of a-mono-derivatives, by the action of sodium and alkyl iodides, and, in general, in those cases where substitution is known to occur in the a-position in other ketones. Lowry has shown (Trans., 1898, 73, 569) that the ordinary a-dihalogen derivatives of camphor contain both halogen atoms in this position, and has suggested that the substituent in this position be termed a- or a'-, according to its space orientation with regard to the camphor nucleus.

A second position in the camphor nucleus is sometimes liable to attack; the hydrogen at this point, known as the B-position, appears, however, to be replaced only through the occurrence of isomeric change in an a-brominated derivative, under the influence of hydrogen bromide. The new position occupied by the bromine atom in these derivatives is not known; it may, however, be mentioned that a similar migration of a bromine atom in an a-brominated ketone has been noticed by Conrad and Schmidt, who found that when bromine is added to ethylic acetoacetate dissolved in carbon bisulphide, the amount of y-bromo-derivative, CHBr CO-CH2 COOEt, produced depended solely on the length of time the initial product,

CH, CO CHBr COOEt,

remained in contact with the hydrogen bromide produced in the reaction (Ber., 1896, 29, 1042). If, then, the change of aa'-dibromocamphor into aß-dibromocamphor is of a similar nature, the latter would be represented as containing the group :CBr CO-CHBr, as

Tiemann has suggested, in explanation of its transformation into campholenic acid under the influence of sodium amalgam.

A third series of derivatives, known as the π-derivatives, is obtained indirectly from the corresponding sulphonic acids which are produced when camphor and its a-halogen derivatives are treated with an excess of sulphuric or chlorosulphonic acid. These appear to contain the substituent in a methyl group (Kipping, Trans., 1896, 69, 920), and may, therefore, be represented by the general formula

To these may possibly be added a new type, represented at present by the newly discovered sulphonic derivatives produced by treating camphor with a mixture of sulphuric and acetic anhydride (Reychler, Bull. Soc. Chim., 1898, [iii], 19, 120); these are not identical with the π-derivatives, but may possibly be a- or ẞ-derivatives. So far, however, no evidence as to their constitution has been adduced.

The sulphonic group, being of an "unsaturated" type similar to that of the carboxyl group, might be expected, under favourable conditions, to render the hydrogen of the contiguous carbon atom replaceable by substituents; this also appeared probable from the fact that, in the aromatic series, the sulphonic group renders the "meta" hydrogen atom directly replaceable, thus being brought into line with those groups which, in the fatty series, lead to the production of a-substitution derivatives. It seemed not impossible, therefore, that, by bromination of a-bromocamphor--sulphonic acid, a dibrominated acid containing the group :CBr SO,H might be produced, and the investigation of the compound could hardly fail to afford further evidence as to the true position of the sulphonic group in the camphor nucleus. On the other hand, it was equally likely that the substitution might take place in the a-position, giving a sulphonic acid of ordinary aa'-dibromocamphor, or in the B-position giving an acid derived from aẞ-dibromocamphor. The question was therefore one of considerable interest, and the present paper contains an account of certain derivatives of the acid in question, and of the methods employed for determining its constitution.

The dibromo-acid was first obtained by Dr. Kipping and the author (Trans., 1897, 71, 19) by acting on a-bromocamphorsulphonic acid with dilute nitric acid, the additional bromine atom having been supplied by the destruction of the larger portion of the monobromoacid. Its true nature, however, was not at first recognised, and the substance was thought to be hydroxydibromocamphorsulphonic acid, C10H14Br2SO; this formula, as we have since shown (Proc., 1898, 14, 159), should in reality be C10H14Br2SO4 + H2O.

5

The method of preparation, however, left much to be desired, as at

least half the original substance is broken down, and the difficulty of isolating the dibromo-acid was increased by the presence of large quantities of oxidation products. Experiments were therefore made with the object of ascertaining if the substance could not be prepared by the direct action of bromine on the monobromo-acid. It was found that, whilst the dry ammonium salt was but slowly attacked by bromine when heated with it in open vessels at temperatures below 100°, a nearly quantitative yield of the dibromo-derivative could be obtained by heating an aqueous solution of the salt with the calculated amount of bromine in sealed tubes at 100°, this method being used throughout.

Having obtained in this manner fairly large quantities of the dibromo-acid, its properties were studied more closely than was previously possible. Its salts were found to be well defined substances as a rule, and in the case of the barium and lead salts, were less soluble than those of the monobromo-acid; the basic lead salt is practically insoluble, whereas that of the monobromo-acid is readily soluble in cold water.

A number of new derivatives have also been examined, including the sulphonic chloride, amide, and piperidide; these, like the sulphonic bromide previously described, may readily be obtained in large crystals. The sulphonic chloride resembles closely in appearance, crystalline form, and mode of development, the sulphonic bromide, and the axial ratios of the two substances are not widely different; both compounds, as well as the sulphonic amide, belong to the orthorhombic system and exhibit the phenomenon of sphenoidal hemihedrism.

In the former paper, it was also shown that, by boiling the aqueous solution of the acid with silver nitrate and evaporating the filtered liquid to dryness, a neutral substance having the formula C10H12Br2SO4, may be obtained; as this is not formed when alkali is used, much attention was paid to the question of the action of alkalis on the various derivatives of the dibromo-acid, and it was found that they exert a reducing influence, producing derivatives of a-monobromocamphorsulphonic acid. This mode of action of aqueous alkali is not often met with in other series, but is very characteristic of the aa'-di-derivatives of camphor; thus, the aa'-dihalogen compounds, and aa'-bromo- or -chloro-nitrocamphors lose halogen readily under the influence of caustic potash or soda, being reduced to the corresponding a-mono-derivatives.

The action of piperidine on the sulphonic chloride is somewhat curious; when the action was restrained by the presence of ether as a diluent, the product, although at first insoluble in cold water, gradually dissolved in it; moreover, a considerable proportion of the product of the reaction is monobromocamphorsulphopiperidide, and

the production of this compound must have been the result of a partial oxidation of the piperidine present. It may be mentioned that an exactly similar change occurs when aa'-dibromocamphor and aa'-bromonitrocamphor are heated with piperidine, one atom of bromine in these substances being replaced by hydrogen.

When dibromocamphorsulphonic bromide is heated, decomposition occurs, sulphur dioxide being evolved and a new tribromocamphor formed; this melts at a temperature far below the melting point of an-dibromocamphor, and its properties at once recall those of ordinary aa'-dibromocamphor. Thus it is very stable towards strong nitric acid, but in presence of silver may be readily oxidised by the dilute acid; its alcoholic solution, when mixed with alcoholic silver nitrate, at once begins to deposit silver bromide; it is rapidly attacked by alcoholic potash, one bromine atom being lost and an-dibromocamphor produced; when, moreover, an-dibromocamphor is heated with bromine in a closed tube, the new tribromocamphor is the only product. Its relationship to an-dibromocamphor is therefore doubtless of the same nature as that of aa'-dibromocamphor to a-monobromocamphor. The substance is, therefore, aa'-tribromocamphor,

CBr2

its structure being probably CH,Br.CH11<co2.

11

The following scheme represents, in an abbreviated manner, the relationships which have been established between the various products:

Purified ammonium dibromocamphorsulphonate (20 grams) is placed in a thick-walled glass tube, together with 50 c.c. of water and a small, thin-walled closed tube containing the requisite amount of bromine

(10 grams), the latter being broken after the outer tube has been. hermetically sealed. The whole is then heated at 110-120° during 3-6 hours, when, as a rule, the colour of the bromine has disappeared; the liquid in the tube either sets, on cooling, to a mass of shining needles, or crystallises when the interior of the opened tube is rubbed with a glass rod. Experiments made in this way, however, are not uniformly successful, as in some cases the liquid becomes black, owing to partial or complete carbonisation; it is difficult to say what precisely determines this effect, as of two tubes having nearly the same dimensions, filled apparently in the same manner and heated side by side, one may behave quite satisfactorily, and the other contain a black liquid. It may be stated, however, that carbonisation may be entirely, or almost entirely, obviated by employing a much larger quantity of water, but in that case crystals are not deposited on cooling, and the subsequent isolation and purification of the product becomes a matter of much greater difficulty.

The crystals consist, for the most part, of dibromocamphorsulphonic acid, contaminated with its ammonium salt and ammonium bromide, and may be separated from the strongly fuming liquid by filtration, preferably through asbestos, washed with a little water, and dried at 100°. The mass is then powdered, transferred to a dry flask, and extracted as rapidly as possible with boiling ethylic acetate, which leaves the ammonium salts practically untouched; the filtered liquid, which quickly becomes dark coloured, should be rapidy evaporated to dryness, and the residual mass recrystallised from water to which animal charcoal has been added.

The acid prepared in this way was shown to be identical with that obtained during the oxidation of ammonium bromocamphorsulphonate with nitric acid, by a direct comparison of the two substances, and by the fact that when its aqueous solution was warmed with silver nitrate, silver bromide was formed, and on filtering from this and evaporating to dryness, the lactone, C10H12Br2SO4, melting at 188-189° was obtained.

The acid crystallises with water, apparently in several different proportions, and some of this water is obstinately retained at temperatures considerably above 100°, and even after fusion, owing to the fact that lower hydrates of high melting point crystallise from the hot liquid. Thus, as obtained by very slow evaporation of the aqueous solution, the crystals melt, somewhat indefinitely, below 100°, the exact temperature depending on the rate of heating; when obtained under other conditions, by rapid crystallisation from hot water, or by allowing the foregoing crystals to effloresce, it melts at 128-133°, giving off water vapour, and solidifies at a slightly higher temperature; further heating causes it to melt at 156-159°,