has, up to the present, been recorded (see Guthzeit, Ber., 1893, 26, 2795; Ruhemann and Sedgwick, Ber., 1895, 28, 825; Errera, Gazzetta, 1897, 27, ii, 393; Ruhemann and Browning, Trans., 1898, 73, 280). On hydrolysis with hydrochloric acid, it loses two mols. of carbon dioxide, and yields 2: 6-dihydroxypyridine (Ruhemann, Trans., 1898, 73, 350; see also Errera, Ber., 1898, 31, 1246).

We find that the ethyl ester of the remaining dicarboxylic acid,

is readily prepared by placing concentrated aqueous ammonia with ethyl propenetetracarboxylate in a bottle, and agitating the mixture from time to time. In the course of 4-5 days, the oil disappears and the vessel contains a reddish solution, together with a yellowish solid. The latter we have not examined, since its amount is very small. The filtrate is concentrated by heating under diminished pressure on the water-bath, and then acidified with hydrochloric acid. This precipitates a solid, which dissolves readily in hot alcohol, but is only sparingly soluble in boiling water, and separates from the hot aqueous solution in colourless prisms, melting at 161-162°. Ferric chloride gives a reddish-violet coloration with aqueous or alcoholic solutions of the substance. On analysis:

0.2100 gave 0.3975 CO, and 0.0993 H2O. C-51.62; H=5.25. 0.2340 115 c.c. moist nitrogen at 14° and 745 mm. N = 5.66. C11H1ON requires C=51.76; H=5·10; N=5.49 per cent.

13

Ethyl 26-dihydroxypyridine-3: 4-dicarboxylate, besides acidic properties, has also those of a base, since it dissolves in concentrated hydrochloric acid with the greatest ease. On boiling this solution for 3 hours, hydrolysis takes place, accompanied by the removal of 1 mol. of carbon dioxide, and there is formed a monocarboxylic acid, which is precipitated by adding water to the acid solution. This acid has been identified as citrazinic acid by direct comparison of its properties and by the following analyses :

0.1465 gave 0.2485 CO, and 0.0405 H2O. C-46.26; H=3.03. 0.2480 18.5 c.c. moist nitrogen at 9° and 769 mm. N=9.08.

CH2ON requires C-46.45; H=3-22; N=9.03 per cent.

The decomposition which ethyl 2: 6-dihydroxypyridine-3: 4-dicarboxylate suffers on heating with hydrochloric acid is in accordance with the behaviour of the corresponding ester of the 3: 5-substituted dicarboxylic acid which, as mentioned before, under the influence of boiling hydrochloric acid, loses 2 mols. of carbon dioxide. These facts show that 2: 6-dihydroxypyridinecarboxylic acids with CO2H groups

in the 3-position readily lose them with elimination of carbon dioxide. In conclusion, it may be stated that further experiments on the lines indicated in this paper are in progress.

GONVILLE AND CAIUS COLLEGE,
CAMBRIDGE.

XXV.-Studies in the Camphane Series. Part I. Nitrocamphane.

By MARTIN ONSLOW FORSTER.

14

In a recent account of the effect produced on camphoroxime by the action of potassium hypobromite (Trans., 1899, 75, 1141), I gave a brief description of the resulting compound, which has the empirical formula C10H16O,NBr. At the time of its discovery, the compound itself was not submitted to close examination, because the chief interest attaching to it lay in the fact that, by a series of apparently simple changes, it may be transformed into isolauronolic acid, C,H1402. The action of potassium hypobromite on oximes has not been studied systematically, but Piloty has shown that bromine converts acetoxime into bromonitrosopropane (Ber., 1898, 31, 452), and when camphoroxime is oxidised with potassium ferricyanide or permanganate, a nitroso-compound is produced. These considerations, coupled with the result of applying Liebermann's test, led to the belief that the compound C10H160,NBr is a bromonitroso-derivative, but it was pointed out that the oxidation of camphoroxime with potassium hypobromite not only involves removal of hydrogen, as in the conversion of acetoxime into bromonitrosopropane, but results also in addition of oxygen, so that the apparent analogy between the two changes is not complete.

In the present paper, evidence is brought forward to show that the product of the action of potassium hypobromite on camphoroxime is a nitro-compound, in spite of the fact that it gives Liebermann's reaction for nitroso-derivatives, and assuming Bredt's structural representation of camphor to be correct, there is every reason to believe that the substance has the constitution expressed by the formula

If the saturated hydrocarbon, C10H18, of which camphor is a ketonic derivative, is spoken of as camphane, the above compound must be

called bromonitrocamphane, and it becomes necessary to adopt a prefix which will show that both substituents are attached to the same carbon atom. Considering that the actual arrangement of the carbon atoms in the camphor molecule is still under discussion, it would be premature to indicate each one by a number, on the principle adopted for derivatives of benzene or naphthalene, but as the existence of the group CO-CH2 in camphor is universally admitted, there can be little objection to using the first two numerals in the formula for camphane as represented by the scheme

This order appears more suitable than the converse because in the future, when the constitution of camphor has been finally decided and the orientation of camphane derivatives requires elaboration, the carbon atom of the methylene group which corresponds to the carbonyl radicle in camphor will become a natural starting point in any system of numbering which may be ultimately adopted. The use of these numerals will obviate confusion with derivatives of camphor, which are distinguished by letters of the Greek alphabet; at the same time, it leaves room for differentiating the possible stereoisomerides of a disubstituted camphane in which the substituents are attached to the same carbon atom.

According to this nomenclature, borneol and bornylamine would be referred to as 1-hydroxycamphane and 1-aminocamphane respectively, and the product of the action of potassium hypobromite on camphoroxime is now called 1: 1-bromonitrocamphane.

When bromonitrocamphane is heated with alcoholic potash, bromine is replaced by hydrogen, and 1-nitrocamphane is produced. From the CH2 2

formula of this compound, CHCH-NO, it will be recognised

14

that, as a secondary nitro-derivative, it should resemble phenylnitromethane in its property of occurring in two modifications. This resemblance has been found to exist, nitrocamphane being obtainable in two solid forms, one of which, representing the normal modification, is stable, melts at 147-148°, and has [a] +4.6°, whilst the pseudomodification, which rapidly changes spontaneously into the normal form, melts at 74°, and has [a] - 95° (approximately). The relation between these two substances corresponds in nearly every respect with that subsisting between the two modifications of phenylnitromethane (Hantzsch and Schultze, Ber., 1896, 29, 699, 2251), which has been further illustrated by the study of similar nitro-compounds (compare Konowaloff, Ber., 1896, 29, 2193; Lowry, Trans., 1898, 73, 986;

1899, 75, 211). The normal isomeride, for instance, is indifferent towards alcoholic ferric chloride, which instantly develops a deep cherry-red coloration with the labile modification; the latter also combines with phenylcarbimide and alcoholic copper acetate, both of which are without action on the stable form. Normal nitrocamphane is not attacked by aqueous sodium carbonate, but dissolves in caustic potash, forming the potassium derivative of the pseudo-modification; the latter isomeride dissolves readily in sodium carbonate, without liberating carbon dioxide, and this gas reprecipitates it from solutions in alkalis. From

14

and CH14

CH, the formulæ, CHCH NO2 14 CH.NO,' which represent a-nitrocamphor and 1-nitrocamphane respectively, it is clear that the substances are closely related; in spite of general similarity in properties, however, they differ from one another in an important characteristic. Solutions of normal nitrocamphane in alcohol or benzene do not exhibit mutarotation, and hitherto the conversion into pseudonitrocamphane has been effected only by the agency of concentrated aqueous potash; the labile modification, liberated from solutions in alkali by addition of acid, changes spontaneously when solid or dissolved, and becomes almost completely transformed into the normal modification. Normal nitrocamphor, on the other hand, changes into the pseudo-form when dissolved in organic solvents, but the transformation affects only about 7 per cent. of the material, when a condition of equilibrium is reached (Lowry, Trans., 1899, 75, 216).

It is noteworthy that the conversion of pseudonitrocamphane into the stable modification is greatly accelerated by the action of light. The diminution in specific rotatory power which is displayed by an alcoholic solution preserved in darkness during three hours, was found to be only two-thirds of that undergone by a portion of the same liquid when exposed to light during the same interval of time; in other words, the period required by a solution of pseudonitrocamphane to reach a given degree of specific rotatory power is about twice as long when the liquid is protected from light as it is when the substance is exposed to this influence. Piperidine and alkalis also exert an accelerating effect on the transformation of the labile compound.

In the present paper, I have represented pseudonitrocamphane by the formula CH18:N(OH):O, this expression being of the type R.CH:N(OH):O, proposed for pseudonitro-compounds by Nef. It is unnecessary to discuss the general question of the advantages which this type offers over that represented by the formula R⚫CH

N.OH

origin

ally suggested by Hantzsch, because the last-named investigator now

employs Nef's formula (Ber., 1899, 32, 575). But its application in the case of pseudonitrocamphane may appear to require some justification because the closely allied substance, pseudonitrocamphor, has been

recently represented by the alternative formula, C.H14

N.OH

(Lowry, Trans., 1899, 75, 212). Two reasons for this course are given by Lowry (Trans., 1898, 73, 995) in the following terms, "first, the extremely high rotatory power of the substance is most readily N.OH in which the a-car

explained by a formula such as CH14

bon atom is asymmetric, rather than by a formula in which this

asymmetry does not occur, as in CH14 C:NO.OH

Secondly, the decomposition of the anhydride of pseudonitrocamphor by heat into camphorquinone and nitrous oxide cannot be explained in any simple way except by the use of a formula such as that already given for the substance, in which one oxygen atom is shown to be attached to the a-carbon atom." These arguments, however, are not convincing. In the first place, the high rotatory power of the substance is explained quite as readily by the existence of the unsaturated linking in the C:N(OH):0 CO as by the asymmetric carbon atom in the alternative formula, because Haller and Muller have shown that

expression CH1

C:CHR

CO

compounds of the type CH14 have a specific rotatory power which, in some cases, is ten times greater than that of camphor (Compt. rend., 1899, 128, 1370; compare also Forster, Trans., 1899, 75, 1149). In the second place, direct attachment of oxygen to the a-carbon atom cannot be claimed in the case of chloronitrocamphor, CH14 C ÇCI.NO2 , yet this substance, like the anhydride of pseudonitrocamphor, yields camphorquinone when heated; these are probably examples of intramolecular oxidation, of which several instances are known, notably the action of sunlight on benzylidene-o-nitroacetophenone, which yields indigo and benzoic acid,

(Engler and Dorant, Ber., 1895, 28, 2497). As evidence of the unsaturated character of pseudonitrocamphane, it may be mentioned that in chloroform solution it decolorises bromine instantaneously, and