148 
Diethylisoanthraflavine, Ci 4 ,H 6 (C 2 H 5 ) 204 , was prepared 
ill the same way as dieth^danthraflavine, which it closely 
resembles. It crystallises from alcohol in long light yellow 
shining needles, soluble in alcohol and ether, more soluble 
in glacial acetic acid and benzol. It fuses at 193 — 194°. 
It dissolves in concentrated sulphuric acid, forming a purple 
solution, the spectrum of which shows two ill-defined ab- 
sorption bands, one in the yellow, the other in the green. 
Isoanthraflavine gives with fuming nitric acid a nitro- 
substitution product similar in its appearance and general 
properties to nitro-antliraflavic acid, but we have been 
unable from want of material to examine it minutely. 
We will conclude with a few remarks on the action of 
caustic alkalies on anthraflavine and isoanthraflavine. 
On a former occasion it was stated by one of us that 
anthraflavine yields by the action of fusing hydrate of pot- 
ash alizarine, and it was this supposed convertibility into 
alizarine which led Liebermann and others to the conclusion 
that anthraflavine was identical with monoxyanthraquinone. 
We are now in a position to assert with confidence that the 
product of the action of alkalies is not alizarine. On 
repeating the experiment on a larger scale by fusing the 
substance with caustic potash in a silver basin we obtained 
a substance which after being freed from impurities crystal- 
lised from alcohol in orange-coloured needles strongly resem- 
bling but certainly not identical with alizarine. On heating 
it yields a sublimate in needles very similar to sublimed 
alizarine. The solution in alkalies is however devoid of the 
violet tint characteristic of alizarine, and on dilution appears 
distinctly red. The spectrum of the solution shows two 
absorption bands similar to those of alizarine solution, but 
these bands, according to the determination kindly under- 
taken for us by Dr. Schuster, lie further away from the red 
end than the bands of alizarine. It appears probable that this 
body may turn out to be an isomeride of purpurine, resem- 
