of chlorine-water, and heated in a bottle at 100° for seven hours. On testing a portion of the product, it was found to consist of unaltered substance. The aqueous liquid was then removed at the pump, and the substance transferred to a larger bottle, 300 c.c. chlorine-water added, and then heated at 100°, as usual, for six hours. On opening the bottle, it was found that there was considerable pressure; oxides of nitrogen were given off, and there was a strong odour of chloropicrin. The portion soluble in potassium carbonate solution consisted, however, entirely of unaltered dichlororthonitrophenol salt.

Nitration of Orthochloroparabromophenol.

Orthochloroparabromophenol was prepared as stated in a previous portion of the paper by chlorinating parabromophenol with SO2Cl2 ; the parabromophenol employed was pure, but the chlorination product was not examined, further than it was found to consist of colourless needles. The experiment to be described was carried out as follows: 38 grams of pure parabromophenol was fused in a flask, and after cooling somewhat, but not to the solidifying point, 29 grams of SO2Cl, was slowly run in with constant shaking. It was then allowed to stand, after which it was first freed from acid, and afterwards dissolved in about 200 c.c. glacial acetic acid, and the solution warmed with a slight excess of nitric acid (sp. gr. 1·42). After standing, it was thrown into water, and the precipitated nitro-compound filtered off at once: when dry it weighed 35 grams. From the mother-liquor, a further quantity of nitro-compound precipitated on standing, which was found to give a yellow potassium salt; the latter, on recrystallisation, was obtained in long, yellow needles. Analysis:

0.2076 gram lost 00126 gram at 120° and gave 00612 gram K2SO.

From the above numbers as well as from the melting point of the phenol, 137° with decomposition, there is no doubt whatever that this compound was potassium chlorobromoparanitrophenol (compare this vol., p. 57).

The main portion, which was also converted into potassium salt, was found to consist chiefly of orthochloroparabromorthonitrophenol melting at 114°, but orthochlorodinitrophenol melting at 110° was also separated. The following are the analytical results :

Potassium Orthochloroparabromorthonitrophenol.

I. 0-2319 gram dried at 100° gave 0.0717 gram K2SO.

II. 0.1963

0.0598

Potassium Orthochlorodinitrophenol.

0.2746 gram lost 00150 gram at 130° and gave 0.0870 gram

Action of Bromine on Parachlororthonitrophenol (correction).

In my former work the mistake to which I have already alluded must have occurred somewhat in the following manner :-' -The first portions of parachlororthonitrophenol which I used I took the trouble to thoroughly purify by conversion into a salt and recrystallisation (loc. cit., 787), but, in some of the subsequent experiments, I made use of a nitro-compound, which had been obtained from crude monochlorophenol, and which was consequently contaminated with orthochlorophenol, and it would seem that in the experiments in which I decided that isomeric change had taken place, the chloronitrocompound must have consisted almost entirely of orthochlororthonitrophenol, since the substance I obtained by the action of bromine. at a high temperature undoubtedly consisted chiefly of orthochloroparabromorthonitrophenol.

In repeating this work, 10 grams of parachlororthonitrophenol, which had been very carefully prepared from solid parachlorophenol and had the correct melting point 87°, was dissolved in glacial acetic acid, and the solution heated to boiling; 9.5 grams of bromine (1 mol.) was then weighed out and slowly added to the solution; when half the bromine had been run in, the solution was again heated to boiling, and the remainder of the bromine added slowly as before; after the addition of all the bromine, the mixture was again heated, and finally allowed to cool and poured into a large bulk of water. The crude substance, after crystallisation from spirit, melted at 123-124°. On conversion into potassium salt, the latter was first of

all obtained in dark-coloured plates, but these on recrystallisation from water had the form of delicate red needles, which gave on analysis K = 13.67 per cent.-theory requires 13:43 per cent.-and on decomposition with acid, a nitro-compound, which, when crystallised from spirit, melted at 125° (compare this vol., p. 60).

The calcium salt was obtained in small, dark-red, glistening plates. Analysis:

I. 0-3646 gram lost 0.0424 gram at 100-180°, and gram CaSO..

II. 0-3555 gram gave 0·0765 gram CaSO4.

0.2869 silver precipitate.

III. 0.2667

gave 0.0789

The various specimens of this calcium salt which gave these analytical numbers were obtained from different portions of a solution, so that there is no doubt as to the homogeneity of the same. On decomposing a portion with acid, and crystallising the resulting compound from spirit, it was found to melt at 125°.

Nitration of Parachlororthobromorthonitrophenol.

In order that the constitution of the foregoing compound might be further confirmed, 4 grams of the crude substance, m. p. 123–124° (see above), was dissolved in hot glacial acetic acid, and a slight excess of nitric acid (sp. gr. 1·42) added. The solution was again heated to boiling. On cooling, needle-shaped crystals separated. The compound was converted into the potassium salt, which, after recrystallisation from water, was obtained in long, lightish-red needles. Analysis:

On decomposition with acid, a nitro-compound was obtained which melted at 78-79°. The melting point of parachlorodinitrophenol is

80.5°.

VOL. LV.

2 T

The constitution of the compound obtained by brominating parachlororthonitrophenol at a high temperature is, therefore, parachlororthobromorthonitrophenol, or, in other words, it is a normal product.

Nitration of Parabromorthochlororthonitrophenol.

This substance was nitrated in a manner similar to the foregoing isomeride. The potassium salt of the product was obtained in bright yellow needles. Analysis:

:

The phenol from this melted at 110°, which is the correct melting point for orthochlororthoparadinitrophenol,

My experiments serve to show that there is a tendency for bromoderivatives of orthonitrophenol to be converted into those of paranitrophenol when heated with bromine and water, although they throw no light on the exact manner in which the change takes place. It might appear that the production of paraquinonederivatives from those of orthonitrophenol necessitated the transference of the NO-group from the ortho- to the para-position, but this view is excluded by the fact that whereas dichlororthonitrophenol is readily converted into a paraquinone-derivative when heated with bromine and water, dichloroparaphenol is certainly not produced in the reaction.

LVII.-Benzyl-derivatives of the Phenylenediamines.

By RAPHAEL MELDOLA, F.R.S., and J. H. CoSTE.

In a paper published last year by one of the authors and E. H. R. Salmon (Trans., 1888, 774) it was shown that para- and metanitraniline could be readily benzylated by heating with benzyl chloride in the presence of aqueous caustic soda solution. As the benzylderivatives of the corresponding diamines have not hitherto received much attention, we have thought it desirable to prepare these compounds and some of their derivatives in order to study the effect produced upon the properties of these derivatives by the presence in

the molecule of a heavy positive radicle such as benzyl. Dibenzylparaphenylenediamine has been described in a paper published by Matzudaira (Ber., 1887, 1614).

BENZYLPARAPHENYLENEDIAMINE.

Benzylparanitraniline was reduced by boiling the alcoholic solution with tin and hydrochloric acid; the alcohol was distilled off and the tin precipitated by sulphuretted hydrogen. On evaporating the solution, white, scaly crystals separated out, and after purification by dissolving in water, filtering, and again evaporating in the presence of excess of hydrochloric acid, the salt was obtained in the form of lustrous, white scales which proved to be the dihydrochloride: 0.1538 gram burnt with lime gave 01639 gram AgCl. 0.1135 gram gave 9.7 c.c. moist N at 9° C. and 764-3 mm. bar.

:

The dihydrochloride is readily soluble in hot and cold water, and is thrown out in the crystalline form by the addition of hydrochloric acid. The salt is almost insoluble in cold alcohol. The aqueous solution, not containing an excess of acid, has the following properties::

Ferric chloride produces a green colour at first but this soon changes into red, the latter colour becoming more pronounced on heating the solution and an odour of benzaldehyde being at the same time evolved. Potassium chromate produces a deep bluish-green precipitate which on boiling becomes red and gives off benzaldehyde. Bromine-water produces a green colour which changes to red on standing. When the solution is saturated with sulphuretted hydrogen, the addition of ferric chloride causes a blue precipitate of a colouring matter belonging to the methylene-blue group.

The free base is precipitated on adding alkali to the aqueous solution of the dihydrochloride in the form of an ochreous precipitate, which under the microscope is seen to consist of an emulsion of oily globules. On extracting the oil by agitating with ether, and allowing the ethereal solution to evaporate, the base separates as an oil which solidifies to a crystalline mass on standing for some hours. The crystals consist of moss-like aggregations having a melting point of 30°. The base soon becomes red by oxidation on exposure to the air.