oxide; for 66 mols. nitrite there were dissolved 10 mols. anhydrous normal hydroximidosulphate. Except for the salt being in beautiful, asbestos-like fibres, there was nothing to distinguish it from the jelly and the wax-like metasulphazate, which therefore we do not hesitate to class as a nitrito-hydroximidosulphate.

Basic sulphazotate, which Fremy considers next, has been shown by us already (loc. cit) to be the 5/6 normal hydroximidosulphate, and not the salt of a distinct acid, the sulphazotic. It is liable to contain a small excess of potassium when crystallised from a strongly alkaline solution. A solution of the normal salt readily deposits it, as does also that of the nitrite compound of the normal salt.

Neutral sulphazotate was shown by Raschig to be the 2/3 normal hydroximidosulphate. Fremy distinguished the potassium sulphazotates from the salts previously described by him by their ability to form other sulphazotates by double decomposition. Fremy's analytical results in the case of the neutral sulphazotate are hopelessly out of accord with its constitution and properties, although those for the basic sulphazotate are satisfactory enough.

Sulphazidate, produced by the hydrolysis of the sulphazotate, is hydroxyamidosulphate (Claus). Sulphazilate and metasulphazilate, oxidation products of the sulphazotate, are ON(SOK), and ON(SO,K), and have been studied by Claus, Raschig, and Hantzsch.

Metasulphazotate.-Sometimes Fremy isolated a salt in the form of spangles (paillettes), in appearance like minute crystals of basic sulphazotate, but differing from them in not being hard. This salt he named, therefore, metasulphazotate. According to him, it can also be obtained by mixing (hot) solutions of the (basic) sulphazotate and sulphazite. It is very soluble in water, very alkaline, and unstable unless the water contains alkali. In pure water, it becomes basic sulphazotate and sulphazite again. It shows the greatest analogy to metasulphazate, and is distinguished from basic sulphazotate in the same way. It may be a compound of basic sulphazotate and sulphazite. So far Fremy. It will be evident that there is nothing in its history or properties to distinguish it, except its occurring in the form of sparkling particles, and even that can be met with in the basic sulphazotate suddenly precipitated; we have also obtained other of the sulphazotised salts in what may be called spangles, although not this particular salt. In the preceding paper (p. 434), we have described an impure form of nitrito-2/3 normal hydroximidosulphate, obtained by dissolving the 5/6 normal salt in a hot concentrated solution of nitrite, but still not so concentrated as to give the nitrito-5/6 normal double salt. This preparation is lustreless while in its mother liquor, but when dried on a tile has a fine silvery lustre. thas, when dried in a desiccator,

VOL. LXXVII.

I I

exactly the composition of Fremy's metasulphazotate, and is much less alkaline than the metasulphazate. It may be formulated as K ̧NS2O+9(KNO„K„HNSO„,1†H2O).

2

Sulphammonate and sulphamidate are respectively nitrilosulphate and imidosulphate (Berglund).

XLV. Camphonic, Homocamphoronic, and

Camphononic Acids.

By ARTHUR LAPWORTH and EDGAR M. CHAPMAN, Salter's Fellow in the Research Laboratory of the Pharmaceutical Society.

In a recent paper by one of us (Lapworth, Trans., 1899, 75, 1134), it was shown that when a-dibromocamphor is warmed with moist silver or mercurous salts, about 15 per cent. of the substance is converted into bromocamphorenic acid, an unsaturated closed-chain compound, first prepared by Forster (Trans., 1896, 69, 46), who obtained it from the same source, but by a less direct method. The observation that the acid could be obtained by the new process was thought to be of much value, as it was thus rendered certain that the relationship subsisting between the structure of the acid and that of the parent compound must be of a comparatively simple character, and all ambiguity concerning the relative position of the bromine atoms in the two substances was removed: a point which had previously been involved in considerable doubt.

The results obtained during an investigation of homocamphoronic acid (Lapworth and Chapman, Trans., 1899, 75, 990), the oxidation product of bromocamphorenic acid, show conclusively that the latter substance contains the complex

It was thought highly probable also that bromocamphorenic acid contained the group CBr:CH as part of the closed chain, and the evidence bearing upon this point has already been discussed in detail (Lapworth, loc. cit.). Furthermore, the assumption was made that, of

the pair of carbon atoms united by the ethylene linking, the one to which the bromine is attached is the nearer to the carboxyl group, but as the reasons for this supposition were not explicitly stated, they may be briefly indicated before entering into the evidence we have subsequently obtained.

The constitution of the oxidation product of bromocamphorenic acid shows that the latter must be either a yd-, or a de-unsaturated acid. The formation of a lactone ring would occur most readily at the y-position in the former and at the 8-position in the latter, that is to say, at the carbon atom which is nearest to the carboxyl group.

The two possible modes of representing bromocamphorenic acid, with regard to the relative positions of the bromine atom and the carboxyl group, may be thus expressed :

•CH:CBr ·CO2H, and CBr:CH

and the corresponding formula for dibromocampholide, its bromination product, are therefore,

Homocamphorenic acid, however, can be converted into a lactone by treating it with sulphuric acid, when it undergoes the isomeric change characteristic of this type of acid, yielding a-monobromocampholide, the constitution of which may also be written in two different ways, corresponding with the alternative formulæ for bromocamphorenic acid, thus:

•CH2 C

2

V.

COO, and CHBг.CH

VI.

The two lactones, dibromocampholide and a-monobromocampholide, however, exhibit certain noteworthy differences; thus the former is reduced with the greatest ease to bromocamphorenic acid, whilst the latter, under far more energetic treatment, remains quite unchanged. The bromine atom of a-monobromocampholide thus has properties entirely different from those of one of the two atoms in dibromocampholide, and it is probably therefore in a different position; since, however, its position is necessarily identical with that of the other bromine atom of the latter, it follows that the two bromine atoms in dibromocampholide are probably attached to different carbon atoms, and the formulæ marked III and V are clearly in accordance with this conclusion.

Further evidence in favour of formula I is afforded by the fact that when camphorenic acid itself is brominated it yields B-monobromo

....

campholide, a lactone isomeric with a-monobromocampholide. On the basis of formula I, the constitution of this substance may be represented as ·CHBr CH.............. COO, whilst if formula II be employed, the formula of the two lactones are identical, unless it be assumed that one is a y- and the other a d-lactone, an assumption which it seems impossible to justify.

The foregoing mode of reasoning, however, did not appear to be altogether unassailable, and in order to obtain unequivocal evidence on this point, we have studied the behaviour of a-monobromocampholide towards hydrolytic agents, and we have again to express our thanks to Dr. Forster who kindly consented to our carrying on investigations in this direction. The point appeared particularly interesting, as the above conclusions lead to the inference that the formation of dibromocampholide by the action of fuming nitric acid on a-dibromocamphor is not merely one of oxidation, but is due to initial hydrolysis with formation of bromocamphorenic acid, which is then attacked by the bromine liberated during the reaction.

In a substance having the structure V above given for a-monobromocampholide, hydrolysis of the lactone ring would be expected to result in the immediate formation of a ketonic acid in accordance with the change thus expressed:

•CH, CBr

COQ + 2KOH=CH, CBr···~CO2K+KBr+ KOH =

он

·CH2· CO· · · · · · CO2K + KBr,

whilst if formula VI were the true one, a dihydroxy-compound would be the final product, or possibly an acid having a ring of the oxide type:

•CHBr⚫CH

•CH(OH).CH(OH)······CO2K+KBr

CH-CHCO,K+KBr

Forster boiled a-monobromocampholide with baryta water and obtained a barium salt of the formula (C10H15O3)2Ba; he describes the appearance, solubility, and melting point of the acid he obtained from it, but does not state that the acid was analysed or closely investigated (loc. cit.). We have carried out the hydrolysis with potash and find that the acid has the formula C10H1603, as he supposed, and exhibits all the properties of a y- or 8-ketonic acid. Thus it gives a crystalline oxime and semicarbazone, and unites with hydrogen cyanide; it also gives a phenylhydrazone and p-bromophenylhydrazone, but these two derivatives, curiously enough, show no tendency whatever to assume a crystalline form, even when prepared from the pure acid and purified with extreme care; they give,

however, crystalline sodium salts, analyses of which indicate that they are derived from the respective hydrazones.

We think it advisable to distinguish this acid by a definite name, and propose to use the term camphonic acid, as the substance is isomeric with pinonic acid. Its constitution is doubtless very similar to that of camphononic acid, and, in accordance with the views already expressed by one of us (Lapworth, loc. cit), will probably be represented by the formula:

according as camphononic acid proves to have the structure,

In view of the close resemblance between the probable formulæ of these two acids, we have carried the study of camphonic acid into considerable detail, as we expected that the experience thereby gained would be of great service in the further study of camphononic acid, for the former is not difficult to prepare in quantities of 40-50 grams, whilst the latter can be obtained only in very small amount.

The point which from the first we regarded as worthy of the closest attention, was the behaviour of the ketonic acids towards hydrogen cyanide. In the case of camphononic acid, which, as we have shown, probably differs from camphoric acid only by the presence of the group :CO in place of :CH CO2H, we hoped that it would be possible to effect a partial synthesis of camphoric acid in a very simple manner. It was to be expected that camphononic acid would combine with hydrogen cyanide, and that the hydroxynitrile thus formed would afford camphanic acid in accordance with the following scheme, using Perkin's formula for camphoric acid.

Our preliminary experiments on the subject showed that camphononic acid, if it united with hydrogen cyanide at all, would do so only under exceptional conditions, and we therefore turned our attention to its homologue, in order to ascertain what would probably be the most suitable mode of procedure. It was soon found that camphonic acid, unlike its homologue, combined readily with hydrogen cyanide, the most suitable conditions being those recommended by Haller and