When dissolved in hydrochloric acid and treated with platinic chloride, the platinochloride was obtained in short, thick, goldenyellow needles.

0.4137 gave 0·1005 Pt. Pt = 24.29.

(CNHCl, NH2)2,H,PtCl requires Pt=24.30 per cent.

The contents of the flask, which were strongly alkaline from the excess of sodium carbonate, after filtration and neutralisation with hydrochloric acid, yielded a precipitate of brown, crystalline matter, very soluble in alkalis, reprecipitated by acids, and moderately soluble in the usual organic solvents. This substance is still under examination, as is also the residue of the steam distillation.

II. In order to avoid the rather considerable amount of decomposition which takes place when the separation is carried out as above described, a modification of this plan was adopted in the later experiments. The solid or semi-solid product of the chlorination was warmed, drained by the aid of the filter pump and washed with chloroform, when a considerable quantity of colourless crystals was left behind. This crystalline separation was purified by several recrystallisations from dry chloroform, dried in a vacuum, and kept in a well stoppered bottle.

It was found that when these crystals were dried in the air they gave off hydrogen chloride for some time after removal to a desiccator. This was doubtless due to interaction with the small amount of moisture absorbed from the air during the process of drying, and does not occur when the substance is dried in a desiccator over sulphuric acid, the crystals in the latter case decreasing in weight by an insignificant amount when left for some days in a vacuum over potash. The compound melted sharply at 187-188° (uncorr.), and gave the following numbers on analysis:

0.7575 gave 0.6050 CO2 and 0·0155 H2O. C=21·72; H=0·22. 0.2830 0-8340 AgCl. Cl=72.86.

C10Cl1HN2 requires C=22.22; H=0·18; Cl=72.22 per cent.

11

It is freely soluble in hot, but only sparingly in cold chloroform, carbon tetrachloride, benzene, or pyridine. On treatment with boiling spirit, an interaction takes place, and very beautiful, highly refractive, colourless crystals are obtained which melt at 171-172° (uncorr.), whilst the mother liquor becomes strongly acid, owing to the production of hydrogen chloride. The same change is brought about on contact with water, but the solid is not taken into solution. Glacial acetic acid dissolves it readily on boiling, with copious evolution of hydrogen chloride, and on cooling deposits the crystals melting at

171-172°. These have the composition expressed by the formula CON2CI,HO.

The presence of hydrogen in the parent substance (m. p. 187-188°) was first attributed to the latter being the hydrochloride of a base, but its occurrence with oxygen in the compound now described was so unexpected, considering the methods of preparation, that numerous analyses were made of specimens from different experiments, and in some cases of relatively large amounts of material, before its existence in the molecule was regarded as placed beyond doubt. The following are typical of the numbers obtained:

0.3639 gave 0.3303 CO, and 0.0078 H2O.

0.5790 0.5250 CO2

0.8110 0-7400 CO2

0.2360 0.6310 AgCl.

0.2200 0.5870 AgCl.

0.0127 H2O.

0.0163 H2O.

C=24-75; H=0·23.
C-24.73; H= 0.24.
C=24-89; H=0·22.

Cl = 66.14.

Cl=65.99.

0.3098 15.5 c.c. nitrogen at 18° and 765 mm. N = 5.84.

CONCI, HO requires C = 24.82; H = 0·21; Cl = 65·86; N=5.79 percent.

A determination of the freezing point of a solution in benzene gave the following numbers:

0.2320 gram lowered the freezing point of 19.59 grams of benzene 0.130°; Mol. wt. = 445; calculated for C1N,Cl,HO = 484.

This substance is readily soluble in chloroform, carbon tetrachloride, benzene, or glacial acetic acid, but only sparingly so in alcohol, and is insoluble in water, dilute acids, or alkalis. Strong sulphuric or nitric acid, on warming, dissolves it with decomposition, and in the former case with liberation of hydrogen chloride. From its solution in chloroform, if deposited slowly, it is obtained in lustrous, highly refractive crystals of large size. It can be crystallised from pyridine without decomposition, and therefore is not the hydrochloride of a base. It is not affected by long-continued boiling with water. The evidence at present is insufficient to decide whether the hydrogen and oxygen in the compound are present as hydroxyl. Its behaviour with phosphorus pentachloride has only partly been examined, but it may be stated that pyridine pentachloride was one of the products, thus showing that the pyridine nucleus still remains. The ease with which the parent substance interacts with liquids containing the OH-group points to the presence of this radicle in the new compound, yet, on the other hand, it may be crystallised unchanged from benzoyl chloride or acetyl chloride.

If the crude cake obtained by draining the contents of the chlorination flask is at once crystallised from alcohol, the first crop of

crystals contains another and somewhat similar compound, which, after purification by several recrystallisations from alcohol, glacial acetic acid, or chloroform, melts unchanged at 228° (uncorr.). This substance gave the following numbers on analysis :

0.4115 gave 0.4350 CO2 and 0·110 H2O. C=28.82; H=0·29. 0.2010 required 0.5814 AgNOg. Cl=60-22.

0.1922 gave 0.4635 AgCl. Cl=59.68.

0.2885 17.3 c.c. nitrogen at 14° and 754 mm. N = 6.99. CON2CI,HO requires C = 29.07; H = 0·25; Cl = 60·00; N = 6.78 per cent. This product is isolated in very much smaller quantity than the preceding one. It is dissolved by the same solvents, but to a less extent. It is probably formed by the interaction of water with a substance of the formula C1N2Cl,H, which has not been isolated.

10 2

The authors are not at present prepared to put forward any constitutional formulæ for these compounds, but the investigation is being pursued, and several interesting derivatives have been obtained. This branch of the subject, however, is reserved for a future communication.

The oily filtrate separated from the crystalline products of the chlorination was freed from chloroform and distilled in a current of steam until nothing further was carried over. The residue, which varied between 5 and 10 per cent. of the total weight of material obtained, was in the form of a light, reddish-brown resin very soluble in chloroform or glacial acetic acid, but only sparingly so in alcohol, and insoluble in water. It is very rich in chlorine, but so far no crystalline products have been obtained from it.

The oil volatile in steam, which was obtained from all the experiments in only slightly varying quantities, was treated in the manner described in Part I. (loc. cit.) for the separation of the chloropyridines. It differs very considerably in composition from the product of the action of phosphorus pentachloride on pyridine, as might be expected in view of the difference in the material employed and the conditions under which the chlorination was carried out. The constituents which do not form compounds with mercuric chloride are present in insignificant proportion, whereas in the former case they reached approximately 50 per cent. of the total weight of oil obtained. From this oil, the only crystalline product is pentachloropyridine.

On fractionating the dried oil, separated from the compound with mercuric chloride (loc. cit.), the first portion of the distillate was found to consist of a dichloropyridine not obtained in the previous work. This compound, which is formed in fair amount, was purified by dissolving in hydrochloric acid, pouring off from any undissolved oil, reprecipitating with ammonia, and subsequent recrystallisation from VOL. LXXV.

3 U

dilute alcohol, when it was obtained in lustrous, flexible flat needles which melted at 66-67° (uncorr.). This fusing point agrees with that of the dichloropyridine described by Königs and Geigy (Ber., 1884, 17, 1833) and the identity of the substance was established by the preparation of the mercurichloride (m. p. = 183°, uncorr.) and the platinochloride, the latter of which gave the following numbers on analysis:

0:36 lost, at 100°, 0·019 H2O and gave 0.0945 Pt. H2O=5.27 ; Pt=26.26.

(CNH ̧Cl2)2,H,PtCl + 2H2O requires H2O=4·84; Pt = 26.48 per cent. The second fraction of the distillate, which was small in amount, consisted of an oil which remained liquid in a freezing mixture of ice and salt. The third fraction, constituting 90 per cent. of the weight of the oil, was redistilled, and the portion boiling between 135-137° under 24 mm. pressure was collected separately. This consisted of the tetrachloropyridine melting at 21-22°, and amounted to upwards of 70 per cent. of the total yield of chloropyridines. Large quantities of this substance have now been obtained and the authors hope to be able to work out its constitution.

As in the former series of experiments (Part I.), the yield of pentachloropyridine is a good one. This substance is chiefly found in the last portions of the steam distillate, which were always collected separately, but is invariably present in the oil separated by treatment with mercuric chloride, and was obtained from it as described in the former paper.

UNIVERSITY CHEMICAL LABORATORY,

CAMBRIDGE.

XCVI.-Homocamphoronic and Camphononic Acids.

By ARTHUR LAPWORTH, D.Sc., and EDGAR M. CHAPMAN, Burrough's Scholar in the Research Laboratory of the Pharmaceutical Society of Great Britain.

THE a-mono-derivatives of camphor are, as a rule, very readily converted into camphoric acid on treatment with oxidising agents such as nitric acid or potassium permanganate, and it is generally accepted that the change is merely one of conversion of the group Co into

-COOH
COOH'

molecule.

CHX

unattended by alteration of the structure of the rest of the

The mode of degradation of the B-halogen derivatives of camphor is quite different from that of the a-derivatives, and the study of the reactions of the products obtained from them, in a manner apparently equally simple, leads to conclusions which, at present, are difficult to entirely reconcile with those drawn from the study of the compounds obtained from camphoric acid. Any new observations as to the manner in which the breaking down of the nuclei may occur can hardly fail, therefore, to be of importance, especially when the changes involved occur at fairly low temperatures, and are effected by agents unlikely to produce violent and deep-seated changes.

Between the a-mono- and the a-di-derivatives of camphor, a remarkable difference in stability is manifested. The former, as mentioned above, are readily oxidised, whilst the di-derivatives are attacked by the ordinary oxidising agents only with great difficulty; thus, a-dibromocamphor is not appreciably affected by ordinary nitric acid (sp. gr. 1.42), although it is attacked by hot, fuming nitric acid (sp. gr. 1.50), and by still stronger acid (sp. gr. 1·52) at the ordinary temperature, whilst aa'-nitrobromo- and nitrochloro-camphor are apparently not attacked even by the strongest acid.

This great stability of the a-di-derivatives is of considerable interest, and their oxidation by mild agents might be expected to afford interesting results. Oxidation products of a-dibromocamphor have already been obtained by Kachler and Spitzer (Monatsh., 1883, 4, 554) and by Forster (Trans., 1896, 69, 36), but in both instances use was made of fuming nitric acid, which involved violent action and the consequent possibility of changes too profound to afford a basis for trustworthy conclusions.

Kachler and Spitzer obtained what they concluded to be camphoronic acid and isocamphoronic acid, together with a substance to which they assigned the provisional formula C24H9BгN4012, whilst admitting its possible want of purity. Forster employed somewhat stronger acid, and found that interaction occurred spontaneously; the product obtained, dibromocampholid, was a lactone which, on reduction, yielded bromocamphorenic acid, C10H15BrO2, and finally, camphorenic acid, C10H1602, a monocyclic, unsaturated isomeride of campholenic acid. The nature of the changes by which dibromocampholid is produced is not quite clear, however, and it is even probable that the bromine atoms are no longer both attached to the carbon atom with which they were associated in the parent substance.

In view of the interest attaching to the fact that dibromocamphor does not afford camphoric acid as a normal oxidation product, the authors have made numerous experiments with the substance in the hope that it could be oxidised by a method sufficiently mild to give results which would be trustworthy from a theoretical point of