MR. SCHUNCK ON RUBIAN AND ITS PRODUCTS OF DECOMPOSITION. 451 
rine and the brownish-red produced by verantine. Pure alizarine is not more soluble 
in boiling alum-liquor than in water, as has been repeatedly shown ; it only commu- 
nicates to the liquor a yellow colour, and crystallizes out again on the liquid cooling. 
Verantine is still less soluble in alum-liquor. If this substance be dissolved in caustic 
alkali and be then precipitated with a solution of alum, the precipitate does not dis- 
solve in the least degree, however much alum be added ; it only communicates a slight 
yellow tinge to the liquid. If, however, a mixture of alizarine and verantine be dis- 
solved in caustic alkali and they be then precipitated together by means of a solution 
of alum added in excess, then on boiling the precipitate with the liquid, a bright red 
solution is obtained, and on filtering and allowing to cool, orange-coloured flocks are 
deposited, while the liquid still remains red, but gives a yellow precipitate on the ad- 
dition of acid. By treating the residue with additional quantities of alum-liquor more 
is dissolved with the same colour, and this continues until either the alizarine or the 
verantine, whichever of the two was present in the smallest quantity, is removed. From 
this experiment I am inclined to conclude that alizarine and verantine are capable of 
forming a double compound with alumina soluble in boiling water, and that a mix- 
ture of the two in the proportion in which they exist in this compound, constitutes 
what has been called purpurine. At all events, it follows that alum is not adapted 
as a means of separating the substances derived from madder. The fact of rubianine 
also dissolving in boiling alum-liquor and crystallizing out again on cooling, is an 
additional objection to its use. 
The difl&culty of obtaining pure verantine in sufficient quantity for the purposes of 
analysis, has prevented me from determining its composition with the requisite accu- 
racy. I have however obtained approximations sufficiently near to remove almost all 
dotibts on the question. 
I. 0'3280 grm. verantine, dried at 100° C. and burnt with chromate of lead, gave 
0*7865 carbonic acid and 0*1215 water. 
II. 0*3220 grm. gave 0*7740 carbonic acid and 0*1205 water. 
III. 0*2890 grm. gave 0*6995 carbonic acid and 0*1040 water. 
IV. 0*1255 grm. gave 0*3010 carbonic acid. 
These numbers agree best with the following composition ; — 
Carbon . . 
Eqs. 
. . 14 
84 
Calculated. 
65*11 
I. 
65*39 
II. 
65*55 
III. 
66*01 
Hydrogen . 
. . 5 
5 
3*87 
4*11 
4*15 
3*99 
Oxygen . . 
. . 5 
40 
31*02 
30*50 
30*30 
30*00 
129 
100*00 
100*00 
100*00 
100*00 
The composition here given approaches that of the oxylizaric acid of Debus, who 
obtained in analysing that substance as a mean of his experiments in 100 parts, — 
Carbon 66*40 
Hydrogen 3*82 
Oxygen 29*78 
