( 301 ) 
On neutralisation of the filtrate with sodium carbonate the 1.3.6- 
derivative m. p. 76° is obtained. 
We now have a complete control over the reaction. If we use 
more than the amount of acid indicated, we run the risk of obtaining 
also trinitroderivatives, which impede the purification of the 1-3-4- 
dinitrodiethy laniline. 
When treated with nitric acid D. 1.37, the compound melting at 
95° passes readily into trinitroderivatives. For this purpose 2.5 grams 
are dissolved in about 21 c.c. of that acid at the ordinary tempe¬ 
rature. After standing for about half an hour yellow crystals are 
deposited, with evolution of heat, which are collected immediately. 
Generally a little nitrous acid is formed. 
When carrying out this experiment for the tirst time, I obtained 
as main product a substance melting at 158° l ). On repeating the 
same in order to prepare a larger quantity and using more nitric 
acid I obtained, however, as main product a compound melting 
at 175°. 
Judging from the analytical results, the tirst one is a trinitrodi- 
ethylaniline and the other a trinitromono-compound. 
The cause of the different course of the reaction was soon found 
out: in the second nitration nitric acid had been used prepared by 
diluting yellow 90 0 / 0 j acid, whereas, in the first one the acid D 1.37 
was obtained by diluting colourless acid D 1.40 with water. Owing 
to the nitrous acid present in. the yellow acid, one ethyl-group would 
have been replaced by hydrogen. 
That this explanation is the correct one could be demonstrated by 
dissolving trinitrodiethylaniline m. p. 158° in moderately strong sul¬ 
phuric'acid, and adding to the solution obtained sodium nitrite. On 
pouring the mass into water the derivative melting at 175° separated. 
This was obtained also by oxidation of the diethyl compound with 
chromic anhydride in acetic acid solution. 
We have succeeded also in determining or deducing the structure 
of the compounds described. 
From analogy with the results of the nitration of dimethylaniline 
it may be expected that the structure of the trinitro-compound 
N(C 2 H s )a 
melting at 158° would 
isomeride melting at in 5 . By 
x ) There is also 
small quantity of 
repeated crystallisations from carbontetrachloride the melting point rose to 
21 
Proceedings Royal Acad. Amsterdam. Vol. XII. 
