P. T. Austen—Dinit lib b ls, 47 
_ 
of heat, the speedy browning of the liquid showed that a more 
vital action had taken place. The best results with this method 
wére obtained by long boiling of the dinitrodibrombenzol 
with an aqueous solution of potassa. Owing, however, to the 
slight solubility of the nitro-compound in water, but a small 
amount of the phenol salt was formed. 
I then endeavored to utilize a reaction which I described, 
some time ago,* as taking place between dinitromonobrom- 
benzol and potassium nitrite, when the two substances are 
heated under pressure in presence of dilute alcohol. The for- 
mation of dinitrophenol takes place easily : 
C,H,(NO,),Br+2KNO,=C,H,(NO,),OK+KBr+N,0,(?) 
The resulting oxide or oxides of nitrogen have here the 
beneficial effect of exerting an oxidizing action. The great ob- 
jection to the use of an alkali in forming hydroxy] derivatives 
from nitro-halides, is the tendency it has to affect the nitro- 
groups converting them into azoxy-compounds, By the use 
of potassium nitrite all reducing action is avoided—in fact, 
prevented—by the presence of the nitrogen oxide. The oxides 
react also on the alcohol present, forming aldehyde. 
Some of the f-dinitroparadibrombenzol was dissolved in 
dilute alcohol, and an equal weight of KNO, added. On stand- 
ing, the KNO, extracts the water from the dilute alcohol, form- 
ing an aqueous layer on which the alcoholic solution rests. By 
standing, even without warming, the line of contact between 
these layers solidifies to a crust of red needles. On boiling, 
however, the action is soon complete, and every trace of the 
f-isomer is converted into the phenol. The reaction is pre- 
cisely analogous to the former : 
C,H,(NO,),Br,-} 2KONO=C,H,(NO,),BrOK+KBr-+N,0,(?) 
_ Some of the pure @ was then treated in a similar manner, but 
did not suffer the slightest change. Some samples of KNO, gave 
a slight coloration, but this was owing to the presence of free 
potassium hydrate. They gave no reaction after the potassium 
nitrite had been moistened with a few drops of dilute nitric acid. 
The oil that I take to be a gamma-isomer, also remained un- 
affected. 
_ This method enables me to overcome several bitherto almost 
msurmountable obstacles. The separation and purification of 
the beta from the alpha by crystallization is extremely difficult, 
and from the almost impossible. By direct treatment 
with KNO,, of the mass obtained by nitrition of the dibrom- 
benzol however, all the beta isomer can be extracted in th 
form of phenol salt, leaving only the alpha, a solid substance 
* This Journal, August, 1876. 
