102 REPORTS ON THE STATE OF SCIENCE. 
anhydrous, diffusing to the region containing the nitroamine, of which at 
most one or two crystals (plates or needles) were used. Obviously this 
procedure admits only of the use of reagents which are not changed by 
prolonged contact with phosphorus pentoxide. 
By this means it has been ascertained that hydrogen chloride, bromide 
and iodide, nitric acid, chlorine, or bromine can all act as transforming 
agents. lodine, hydrogen cyanide (1), sulphur dioxide, carbon dioxide, 
and methyl iodide are inactive. 
Quantitative experiments carried out with hydrogen chloride have 
revealed the very interesting fact that there is a limiting concentration 
(partial pressure) below which hydrogen chloride does not initiate the 
molecular rearrangement. In the case of the nitroamine here dealt with 
the partial pressure of the hydrogen chloride must be approximately 
1/30 atmosphere. 
The behaviour of the nitroamine with hydrogen chloride in various 
solvents was investigated, the most instructive results being obtained in 
aqueous solution, The solution of the nitroamine in pure water can 
apparently be preserved indefinitely in a dark room, no trace of colour 
becoming perceptible, notwithstanding the fact that it is easy to perceive 
the colour communicated by 1/10 mgr. of the nitroaniline to 100 c.c. 
of the solution ; that is, to detect the transformation of 0-1 per cent. of 
the nitroamine. The addition of hydrogen chloride causes transforma- 
tion, provided that the acid is above a certain concentration. The 
nitroamine was used at a concentration of V=207 (1 gr. per litre), 
the concentration of the acid being varied from N to N/1000. In all 
the solutions but that last mentioned, in which the molecular ratio 
C,H,;Cl,. NH. NO,HCl was 5/1, the acid was in excess. The presence 
of the nitroaniline can be detected with certainty at the end of ten 
minutes in the N solution, and after longer periods down to a concentra- 
tion of N/200, below which concentration all change ceases. Similar 
results were obtained with other acids, but, keeping the concentration of 
the nitroamine constant, the limiting value of the effective concentrations 
of the acids varies with the strength of the acid. 
In the hope of isolating some compound of the acid and the nitroamine, 
hydrogen chloride, accurately dried, was passed into a solution of the 
nitroamine in petroleum. A colourless compound separated, which was 
resolved by solutions of alkalis into the nitroamine and hydrogen chloride, 
but when kept in a dry atmosphere the compound slowly changed into 
the nitroaniline. 
These remarkable results greatly illuminate the mysterious part played 
in this type of intramolecular change by the transforming agent (or 
catalyst), the necessity for the presence of which was first recognised by 
Armstrong.! 
The deductions to be drawn from these observations may be thus 
summarised :— 
(i) No ‘specific’ reagent is required for any given transformation. 
(ii) The reagent must be capable of combining with the xitroamine. 
In all probability this union is effected directly with the nitrogen of the 
NH group, which thus becomes quinquevalent. The additive product 
in the case of hydrogen chloride has a composition represented by the 
farmula C;H,Cl,.NH.NO,,HCI. 
' Report, 1899, p. 683, and cf. Trans. Chem. Soc., 1900, 77, 1053. 
