place in solutions of two crystalline derivatives of camphorcarboxylic acid (the amide and piperidide), which were first prepared and examined about four years ago by Dr. Glover. The changes of rotatory power which take place in the freshly prepared solutions are extremely complex. In solutions of the piperidide in benzene, a period of induction is followed by two periods of acceleration and two periods of retardation in the rate of change of rotatory power; the changes can only be explained by assuming that three distinct isomeric changes take place, and that a condition of equilibrium is ultimately established between four distinct isomerides.

Measurements of solubility have shown that in the case of each substance a proportion amounting to about one-third persists in the original form when a condition of equilibrium is finally attained to.

In purifying the two substances it was found to be almost impossible to secure homogeneous material until they were separated in the form of measurable crystals from solutions in ethylic acetate. A remarkable morphotropic relationship was detected between the amide

1

An investigation of the equations for two consecutive unimolecular changes has already been published; and a detailed account of the experiments referred to above will be published at an early date.

The Committee ask for reappointment and for a grant of 30l. as in the preceding year.

The Transformation of Aromatic Nitroamines and Allied Substances, and its Relation to Substitution in Benzene Derivatives.-Report of the Committee, consisting of Professor F. S. KIPPING (Chairman), Professor K. J. P. ORTON (Secretary), Dr. S. RUHEMANN, and Dr. J. T. HEWITT.

I. The Conversion of Chloro-, Bromo- and Nitro-aminobenzenes into Substituted Anilines.

In the recent reports of this Committee, the results of work on the transformation of chloro- and bromo-aminobenzenes into halogenanilides, and of nitroaminobenzenes into nitroanilines has been cominunicated. In the case of the former compounds, it has been shown that the transformation is not an intramolecular change but consists

Trans. Chem. Soc., 1910, 97, 2634-2645.

of a primary reaction of the chloroamine and hydrogen chloride, the presence of which is essential, thus:

Ar. NCIAC + HCl Ar. NHAC + Cl2 · + C1Ar.NHAC + HCl.

Whether a true intramolecular change is possible under certain conditions has not yet been discovered, but it must not be supposed that the possibility is excluded.

The conversion of the nitroaminobenzenes into the isomeric nitroanilines offers a very marked contrast.

(i) All acids, and not one specific substance, bring about the transformation. The relative effectiveness of different acids is generally proportional to their activities in other processes. Moreover, when no sidereactions occur, the speed follows an equation of the first order, and, for a monobasic acid, is proportional to the second power of the concentration, at least in aqueous and dilute acetic acid solution.

(ii) Although there is evidence, but of no certain kind, that the nitration of another substance by a nitroamine can occur (for example, of acetanilide or 2: 4-dichloroaniline by s-tribromonitroaminobenzene), under certain narrowly defined conditions, there is nothing corresponding to the remarkable chlorination of one anilide by the chloroamine of another, which has been described. No radicle (ion) or substance which is a powerful nitrating agent appears to be free in the system.

(iii) The solid crystalline nitroamine changes into the nitroaniline, the crystals of the latter apparently growing out of the former in the presence of gaseous hydrogen chloride in a P,O,-dry atmosphere.1

(iv) Although nitrous acid appears during the transformation, and diazonium salts are produced, the presence of urea in the system does not affect in any way the speed or the products of the change.

(v) The nature of the nitroamine and of the catalyst has a very marked effect on the extent of the side-reaction in which the diazonium salt is produced. The maximum amount of diazonium salt is found with 2 4-dichloronitroaminobenzene, much less with the corresponding 2: 4-dibromo compound, and none with 1-methyl-3-bromo-4-nitroaminobenzene.

The nature of the catalyst has a similar influence. When nitric acid is the catalyst no appreciable quantity of diazonium salt can be found. In the presence of perchloric acid, the maximum amount of diazo compound is produced; hydrogen chloride yields less and sulphuric acid still less. The ratio of diazonium salt to nitroaniline for a given catalyst appears to be independent of the concentration of the catalyst or the composition of the medium; thus in the case of hydrogen chloride and 2:4-dichloronitroaminobenzene in various mixtures of acetic acid and water, the ratio, nitroaniline/diazonium salt-3.7/1.

Conclusion. So far as the evidence goes, the provisional conclusion may be drawn that the conversion of nitroaminobenzenes into nitroanilines differs from the conversion of chloroamines, and may probably

1 Reports, 1909.

be regarded as an intramolecular change. But the possibility that under certain conditions a cleavage into aniline and a nitrating substance occurs, at least partly, cannot be excluded. Thus as an instance s-trichloronitroaminobenzene in an environment when transformation is generally rapid, yields largely s-trichloroaniline.

In the substances which we have closely investigated, one orthoposition only is vacant, into which the nitro group can migrate. The migration of the nitro group into the para-position is observed in the conversion of 2: 6-dibromonitroaminobenzene into 2: 6-dibromo-4nitroaniline; this change is, however, accompanied by the formation of about an equal amount of the isomeric 2: 4-dibromo-6-nitroaniline, the bromine atom in the ortho-position being displaced. From a consideration of the relative proportions of o- and p-nitro-anilines and -anilides produced under various conditions, in the transformation of nitroamines, nitration by acetyl nitrate, &c., Holleman' has concluded that the ortho-compound is generally formed by way of the nitroamine, whilst the para-compound is formed by some other process. In the case above cited, however, the p-nitroaniline is undoubtedly obtained from the nitroamine, although perhaps not by a simple intramolecular re-arrangement. Comparison of the two changes shows at least that the conversion to the o-nitroaniline is a far more rapid and easy process.

4

Formation of Nitroaminobenzenes.-Owing to the difficulties of following further the conversion of nitroamines into nitroanilines, we have been led to study the conditions and mechanism of the formation of nitroamines. An excellent way of converting anilines into nitroamines is by treatment in acetic acid solution with a mixture of nitric acid and acetic anhydride. (Orton and Orton and Edwards.") The behaviour of nitric acid offers in this respect a marked contrast to that of other strong acids; they are powerful accelerators of the direct reaction between the anhydride and anilines (Orton and Smith). In order to investigate these highly distinctive reactions more closely a means of determining acetic anhydride in such systems was required. A good method was finally devised (Edwards and Orton') which is based on the following reactions; certain anilines, for example, 2: 4di-chloroaniline, react with acetic anhydride in acetic acid solution, even containing a certain proportion of water, very rapidly and quantitatively; the anilide is extracted from the diluted medium with chloroform and the excess of aniline removed by treatment with an acid; finally the anilide is converted into a chloroamine which can be estimated by titration. In this manner very small quantities of acetic anhydride can be estimated in the presence of acetic acid.

With the aid of this method of analysis, the hydrolysis of acetic anhydride has been studied and remarkable differences between nitric and other acids, in their effect on this process, discovered.

II. Hydrolysis of Acetic Anhydride. (With MARIAN JONES, B.Sc.)

So far the hydrolysis of acetic anhydride has only been investigated in aqueous solution by Menschutkin and Vasilieff, Lumière and Barbier (these investigators used a titrimetric process), and Rivett and Sidgwick (using the change of electric conductivity). With the aid of the method above described, it is possible to follow this reaction in various media, and in the presence of catalysts.

8

(i) Hydrolysis in Aqueous Solution.--The table shows typical results of experiments when three different methods of measurement are used.

On the supposition that the reaction is represented by the equation: Ac,O+H,O=2CH.CO,H, the velocity factor, k

=

the product of the velocity co-efficient, kn, of this reaction of the second order, and the concentration of the water, which is perceptibly constant,

where A is the initial concentration of the anhydride in moles, and x the amount changed in time t (minutes).

(ii) Effect of Medium.-Table II. shows the effect of the composition of the medium on the rate of hydrolysis.

As the acetic acid is diluted, the rate of hydrolysis increases roughly proportionally to the amount of water in the medium. It is remarkable that pure acetic anhydride should be such a curiously unfavourable medium for its own reaction with water.

(iii) Effect of Catalysts. In aqueous solution alkalis are very powerful catalysts of the hydrolysis. Acids have, on the other hand, but a very feeble effect. Such slightly hydrolysed salts as sodium acetate occupy an intermediate position.

In anhydrous media acids produce a great acceleration of the reaction, but the effect diminishes as the proportion of water in the medium decreases.

With regard to the nature and concentration of the acid, it is to be noted that:

(i) In 90 and 95 per cent. acetic acid molecular quantities of acids are equivalent in accelerating effect.

(ii) In 50 per cent. acetic acid equivalent quantities of acids produce equal effects.

(iii) At intermediate compositions of the medium there is no simple relation.

(iv) When the effect can be measured as in 90 per cent. acetic acid, the rate of hydrolysis is found to be proportional to the concentration of the catalyst.

These relations indicate that the unionised acid is the effective catalyst in media containing 90 per cent. acetic acid and upwards, but that the ionised acid is the catalyst in media containing 50 per cent. acetic acid.

Nitric Acid as Catalyst.-In its relation to the hydrolysis of acetic anhydride, nitric acid occupies a unique position. Whilst in 50 per cent. acetic acid its effectiveness is identical with that of other acids; as the proportion of water in the medium decreases, its relative activity steadily falls off, until in glacial acetic acid it is, compared with sulphuric acid, infinitesimal.