points. Petals 5, imbricate, spreading flat, but becoming 
more or less undulate when beginning to close ; obovate or 
obcordate, slightly emarginate, and crenated with numerous 
very small notches at the point, of a delicate light yellow, 
with a white mark near the base: the flower altogether, 
about the size of a large Primrose. Stamens 10, short, every 
other one longest: filaments flat, tapering to a sharp point, 
and remaining persistent: anthers peltate, or joined by then- 
back to the filaments. Styles 5, villous, unequal in length, 
shorter than the stamens. Stigmas capitate, tubercled. 
Capsules inclosed in the persistent, reticulated calyx, each 
one seeded, all of them so firmly attached together and to 
the calyx, that they cannot be easily separated; it should, 
therefore, be all sown together, and the young plants come 
up readily, the roots bursting through the back of the calyx. 
Our drawing of this very rare and curious plant was 
taken at the Nursery of Mr. Colvill, who, last Autumn, re¬ 
ceived both dried roots and seeds of it from Mr. Synnet, 
who collected them in the interior of the Cape; we had been 
long hoping to see the present species growing in our col¬ 
lections, as we knew of its existence by specimens which we 
had seen in the Banksian herbarium, now in the possession 
of Mr. Brown, Gaertner describes the flowers as blue, and 
they certainly do appear more like blue than yellow when 
dried. 
When we received the roots of the plants, which were 
in good preservation, we expected to see them flower much 
sooner than those plants that were raised from seeds sown 
at the same time; in this we were disappointed, as the seed¬ 
ling plants produced their flowers first, though the old plants 
were also in bud at the same time; the soil in which we 
planted them all, was about one half of potsherds, broken 
small, and mixed with an equal portion of loam, peat, and 
sand; in this they are thriving very well, and producing 
abundance of flowers, and we expect that they will ripen 
seeds. M. Decandolle has now placed this genus in Ro¬ 
sacea. 
1. Ripe fruit, showing the persistent calyx: the capsules and persistent filar 
ments much smaller than in G. tenuifolium, and much less rigid. 2. Stamens, 
every other one longest, all connected at the base, and inserted in the calyx. 
3. Germen, terminated by 5 very unequal woolly styles, terminated by largish 
capitate stigmas. 4. The same with 7 Styles, two of which are very small. 
