RUBIACEH. 209 
low-melting point, somewhat resembling a fat in appear- 
After repeated trials in various ways, it was found that 
3 impurity might be removed by means of light petroleum. 
oiling saturated solution of the gardenin in alcohol was 
allowed to cool, and the almost pasty mass of crystals was 
agitated with light petroleum at a temperature of about 30°, 
the clear liquid poured off, and the residue again agitated with 
petroleum, repeating the operation several times, The ‘gar- 
denin was finally purified by alternate crystallization from hot 
benzine in which it is readily soluble, and from alcohol. 
When pure, gardenin forms brilliant deep yellow crystals, 
which melt at 163° to 164°. Dried at 100°, and burnt in a 
current of oxygen, it gave the following results:— ‘ 
I. 0:249 gram. of substance gave 0°567 gram. carbonic 
anhydride, and 0-119 gram. of water. age 
I. 0-202 gram. of substance gave 0°457 gram. carbonic 
ydride, and 0-102 gram. of water. ee 
‘Theory J. . Mean. Fliickiger. = 
| 60 61°86 62°12 61°70 61:91 59°47 
eS 516 631° 860 545° 671 
92° OB as OR aa eae ieee 
97 - 100-00 
Flickiger’s numbers do not agree with these, but as the 
pecimen he analysed had merely been purified by repeated 
-ystallization from spirit, it is not impossible that it was con- — 
minated with traces of the colourless fatty substance men- 
ned above. This is rendered very probable by the much 
lower melting point (155°) which he obtained. It was stated 
n the earlier paper (Stenhouse loc. ctt.), that when gardenia — 
digested with concentrated nitric acid, itis rapidly decom- 
ed, picric acid, but no oxalic acid, being produced. On 
eating the experiment, however, this statement was found = 
be incorrect ; gardenin when boiled with nitric acid, dis- aoe 
with evolution of nitrous fumes, forming: a yellow 
ion, which on evaporation leaves a yellowish "esi 
, on careful 
