potassium hydrosulphide is formed on passage of hydrogen sulphide through a hot solution of K,S,, and that the liberated sulphur, at the moment of its separation, is taken up by unaltered K,S,, forming the higher compound KS10

(3) It is to be noted that in all cases the polysulphide obtained is richer in sulphur than that for which the sulphur added was sufficient. This requires elucidation and two suggestions are made :

(a) All the sulphur added unites selectively with a portion only of the potassium hydrosulphide, forming a higher polysulphide than was designed in the experiment, and leaving potassium hydrosulphide in solution unaltered.

(b) That the formation of the higher polysulphide is due to the action of the current of hydrogen sulphide on the polysulphide first formed (see paragraph 2).

(4) It is evident on considering the results of experiments a, e, έ, n (Series I), that it is not possible to discriminate between (a) the first product obtained on dissolving sulphur in potassium hydrosulphide, and (b) the condensing effect of the current of hydrogen sulphide on the polysulphide first formed.

As a

(5) Attention may now be directed to the experiments of Series II, (see ẞ, y, 8), in which no current of hydrogen sulphide was employed. Turning to the experiment marked 8 in the Table, the intention was to produce KS, for which the sulphur added was sufficient. matter of fact, the product obtained was K4S, and 50 per cent. of the hydrosulphide was found unaltered after solution of the sulphur had been effected. The following equations represent (i) the action expected, (ii) the change which actually took place.

(i) 4KHS+S2 = K4S5+ H2S

(ii) 8KHS+SKS+4KHS + 2H2S.

The validity of the latter equation seems to be established, for the proportion KHS/S on the left hand side was that existing in the experiment 8 (see Table); and, of the three items on the right hand side, two were definitely established, the existence of KS, and the survival unaltered of 50 per cent. of the original hydrosulphide. It would follow then, from the fact that this proportion remains unaltered, that, by natural selection, under the given experimental conditions, the relation of the reacting materials is not that of the latter equation, but the simpler one: 4KHS+S=K4S8+ 2H2S.

(6) As the result of treatment of crystals of KS, and KS, with carbon disulphide, a lower polysulphide, K,S,,10H2O, has been obtained, which resists further removal of sulphur by this solvent.

Furthermore, KS, is a stable polysulphide, mustard-yellow in colour, easily and completely soluble in water, and in it the following relation

VOL. LXXVII,

3 H

exists, KS, S, that is, of the five sulphur atoms, three are in the polysulphide position. There seems, then, good reason for suggesting that the prime product, in the case of potassium polysulphides, is tetrapotassium pentasulphide, KS,, and that the other crystalline polysulphides obtained are solid solutions of sulphur in this substance.

(7) Regard must also be paid to the fact that the polysulphides of ammonium and sodium are of the same degree of complexity, which seems to indicate that the source of this complexity is in the sulphur molecule itself. Now, it has long been held as a fact that sulphur, at temperatures above its boiling point, possesses a vapour density corresponding to a molecular formula of S, and that only as the temperature rises to 860-1040°, does it conform to the type S2.* During the past few years also, certain papers have appeared dealing with the molecular weight of sulphur in solution (compare Paternò and Nasini, Ber., 1888, 21, 2153; Beckmann, Zeit. physikal. Chem., 1890, 5, 76; Hertz, ibid., 1890, 6, 358; Guglielmo, Real. Accad. Linc., 1892, ii, 210; Orndorff and Terrasse, Amer. Chem. J., 1895, 18, 173), and as a result of this work, the molecule of sulphur is stated to exist in solution as S, Sg, or Sg.

Now the simplest expressions that can be written involving the action of such molecular groupings, and limited by the observed experimental behaviour of potassium hydrosulphide, are as follows: (1) 8KHS+S2K4S, +4H2S. (2) 4KHS+S= KS, +2H2S. (3) 4KHS+S KS10+2H,S. (4) 4KHS+S, K,S11+2H,S. Of these reactions, (2) and (3) have been already obtained, but for (1) and (4) the experimental conditions are not yet known.

=

= =

11

A small portion of this work was carried out at the Royal Naval College, Greenwich, with the aid of Dr. W. J. McKerrow, and the remainder in the Davy-Faraday Research Laboratory of the Royal Institution. The author's thanks are given to the Managers of the Royal Institution for this privilege, and to Dr. A. Scott for kindly criticism and advice received during the course of the investigation.

THE DAVY-FARADAY RESEARCH LABORATORY

OF THE ROYAL INSTITUTION.

* Biltz (Bcr., 1888, 21, 2013) does not, however, consider that the existence of gaseous molecules S can be established. He finds the vapour density varies gradually with the temperature.

LXVI.-Chlorine Derivatives of Pyridine. Part VI. Orientation of some Chloraminopyridines.

By W. J. SELL, M.A., F.R.S., and F. W. DOOTSON, M.A. DURING the progress of the work an account of which has been given in this series of papers, a number of chloraminopyridines have been isolated and described, the structural formulæ of which are but very incompletely known. These compounds, so far as they have been examined, are remarkably stable, and although the fusing points in some cases do not widely differ yet such considerable variations are exhibited in their reactions that identification is rendered comparatively easy. The compounds dealt with are the trichloro- and tetrachloro-aminopyridines and as a basis recourse is had to three substances whose structural formulæ are known beyond doubt. These are (i) tetrachloro2-aminopyridine (Trans., 1898, 73, 779; 1900, 77, 235), (ii) tetrachloro-4-aminopyridine (Amer. Chem. J., 1886, 6, 377; Trans., 1898, 73, 779; 1899, 75, 981), (iii) 3:4:5:6-tetrachloropyridine (Trans., 1900, 77, 2), and are thus represented :

In part III of this series of papers (Trans., 1899, 75, 980) a trichloraminopyridine was described which resulted from the interaction of sodium carbonate with a compound containing two pyridine nuclei. Since this trichloraminopyridine on treatment with phosphorus pentachloride yields a tetrachloraminopyridine identical with I, it follows that the amino-group in the original compound occupies relatively the 2-position and further that in the parent complex the nitrogen of one pyridine nucleus is united to the carbon atom in the 2-position, and not in the 3-position of the other as originally depicted (loc. cit.). It is here further shown that the hydrogen atom in this trichloraminopyridine occupies the 6-position by first converting the amino-group into a hydroxyl group in the usual manner and then heating the product at 182° in a sealed tube with phosphorus pentachloride. The result was a theoretical yield of 3:4:5:6-tetrachloropyridine. These changes may be represented thus:

Similar work has been done in the case of the trichloraminopyridine derived from tetrachloroisonicotinic acid by heating the latter with ammonia (Trans., 1897, 71, 1083). From its genesis, the position of the hydrogen atom in this compound is beyond doubt, but with regard to that of the amino-group no experimental evidence was offered, it being merely suggested that from the known mobility of the chlorine atom occupying the position 2, the constitution of the substance was probably represented by the formula:

On heating a quantity of the trichloraminopyridine with phosphorus pentachloride, it was found that a theoretical yield of the tetrachloraminopyridine (m. p. 174-175°) represented by formula I was ob tained, thus confirming the original conjecture with regard to it.

By the action of ammonia on pentachloropyridine (Trans., 1898, 73, 777) two tetrachloraminopyridines have been isolated one of which has been shown to have the constitution represented by formula II, the other, melting at 174-175°, is identical with I, a structure pointed out at the time as probable, but for which experimental evidence was wanting. From this compound a tetrachlorohydroxypyridine was obtained (loc. cit.) whose constitution VII is thus established. The two tetrachloraminopyridines mentioned above, on further treatment with ammonia and at a higher temperature, both yield the same trichlorodiaminopyridine (loc. cit.) which must therefore be represented by formula VIII. These derivatives are thus represented :

It may here be pointed out that the orientation of the compounds represented by formulæ IV and VI reduces the trichloraminopyridines whose constitutional formulæ are still unknown to two in number and in these cases the hydrogen atom must occupy either the 3- or the 5-position. One of these compounds, melting at 144-145°, is mentioned by Anschütz as resulting from the action of phosphorus pentachloride on B-hydroxyglutaramide (Richter's Organische Chemie, Ed. 1898, ii, 573; see also Trans., 1900, 77, 235).

EXPERIMENTAL.

Tetrachloro-2-aminopyridine from 3:4:5-Trichloro-2-aminopyridine.

The trichloraminopyridine (m. p. 159–160°) derived from the compound containing two pyridine nuclei (Trans., 1899, 75, 980) by distillation with sodium carbonate solution, was heated for 4 hours in a sealed tube at 220-225° with a slight excess of phosphorus pentachloride, the cold mixture treated with water and distilled with steam, when a quantitative yield of the tetrachloraminopyridine melting at 174-175° (uncorr.) was obtained. On analysis:

0.1481 gave 14.9 c.c. nitrogen at 19° and 772 mm. N = 11.75.
0.1080
0-267 AgCl. Cl=61.30.

CH,N,Cl, requires N=12.06; Cl=61.14 per cent.

The properties of the compound thus obtained agree with those of tetrachloro-2-aminopyridine (Trans., 1900, 77, 236).

Conversion of 3:4:5-Trichloro-2-aminopyridine into 3:4: 5-Trichloro2-hydroxypyridine.

Five grams of the compound were dissolved by the aid of a gentle heat in 80 c.c. of sulphuric acid containing approximately 80 per cent. of acid. After cooling, rather more than the theoretical quantity of potassium nitrite was added in small quantities at a time, the mixture heated on the water-bath for half an hour, with occasional agitation, and then, after cooling, poured into 500 c.c. of water. The new compound separated in flocks of fine needles, was filtered off, dissolved in dilute ammonia, to separate any traces of unaltered amino-compound, and again precipitated by the addition of hydrochloric acid.

The substance is readily soluble in alcohol, acetic acid, and most organic solvents. When recrystallised from boiling water, it separates on cooling, either in long, filamentous needles, or in flat needles which are quite colourless. On analysis:

N=7.04.

0.527 gave 32 c.c. nitrogen at 15° and 752 mm. CH2OCI, requires N=7.05 per cent.

When dissolved in warm, dilute ammonia, the ammonium salt, which is comparatively sparingly soluble in water, separates out on cooling in colourless, satiny, micaceous plates, which rapidly lose ammonia on exposure to air and become opaque. The potassium and sodium salts are much more soluble. The silver salt is thrown down as a jelly by precipitation from the ammonium salt.