MR. SCHUNCK ON RUBIAN AND ITS PRODUCTS OF DECOMPOSITION. 105 
incidentally, that in making this experiment it is necessary to treat the verantine 
with a little dilute nitric acid, in order to destroy the alizarine which usually accom- 
panies it, and then to remove the acid by washing with water before employing it, 
and that it generally succeeds best when a large excess of alizarine is used. 
To this synthetical proof of the opinion here advocated, I will now add a few 
analytical ones. I will show, in the first place, that purpurine prepared in the manner 
mentioned by Wolff and Strecker is a substance of very variable composition, but 
that the variations in its composition may be easily explained by supposing it to 
consist of alizarine and verantine in different proportions; and secondly, that by 
treatment with nitric acid purpurine yields unchanged verantine and an acid, which 
is identical with that formed by the action of nitric acid on alizarine. 
In the course of my investigation I obtained at the termination of the process for 
separating the products of the fermentation of rubian, an alcoholic liquid, from which 
the verantine had been deposited, but still containing a substance, which, from its 
solubility in alum liquor, would by most cbemists be called purpurine. This 
substance was precipitated from the solution with water. Its colour was brownish- 
yellow. It was treated with boiling alum liquor, to which it communicated a bright 
red colour. The liquid was filtered boiling hot, and deposited on cooling a quantity 
of red flocks. The residue was treated with fresh quantities of alum liquor, until 
on cooling very few flocks separated. A great proportion of the substance employed 
remained undissolved. The flocks deposited from the alum liquor were collected 
on a filter, and washed with water in order to remove all the alum. After drying they 
formed a dark reddish-brown powder, which was almost entirely soluble in alcohol. 
The alcoholic solution left on evaporation a bright red mass, in which no trace of 
anything crystalline was discernible. Its analysis gave the following results: — • 
0‘4820 grrn. gave T1240 carbonic acid and 0*1800 water. 
0*8060 grrn. left on being incinerated 0*0310 alumina=3*84 per cent. 
After making the proper correction for the alumina, the quantity of which, in rela- 
lation to that of the other constituents, seems to be indefinite, these numbers corre- 
spond in 100 parts to — 
Carbon 66*13 
Hydrogen 4*31 
Oxygen 29*56 
This composition does not differ very widely from that given for oxylizaric acid 
by Debus. 
Now the formula H^o 0, 9 = 0,4 04 - 4 - 30,4 O^ requires in 100 parts — 
Carbon 66*14 
Hydrogen 3*93 
Oxygen 29*93 
