times the length of the tube, and mouth of the tube | 
smooth or nearly so. The throat of fulgidum, crocatum — 
and -pulverulentum appears to be always smooth; that of — 
rutilum, which is distinguishable from fulgidum only by | 
lesser size‘and a shade of colour, has mostly a little excres- _ 
cence at the foot of the upper petal. We have another — 
bulb from Rio, which in bulb, foliage, and habit, ap- 
proaches to fulgidum, with a very pale flower, formed like — 
that of crocatum, having one or two points distinguishable 
on each side of the throat, but more faintly than in sub- 
barbatum, to, which it approximates in no other respect. 
In fuigidum, the filaments are seen through an aperture 
formed by a ventricose curve of the lower petal, which in 
subbarbatum is straight. " | 
We have amended the generic character of Hipreastrum — 
so as to admit reticulatum and striatifolium, abandoning 
the genus Cogureia, which was only separated by the © 
difference of rounder seed: without a hollow foliaceous — 
margin, concerning which separation doubts were ex- — 
pressed in our Appendix. It appears that they breed — 
indiscriminately with any species of Hippeastrum, and the — 
pollen of the mules so obtained is fertile ; and, on further 
examination, it seems that there is no fundamental differ- 
ence in the structure of the seed, but. that the hollow 
margin, which remains unfilled by albumen in the species _ 
which have seeds more numerous and flattened, is com- 
pletely filled in those which produce few and larger seeds. 
A similar habit was observed in the seed of a Brazilian 
Bignonia, now growing in the Spofforth collection, which 
was almost square and not winged, the hollow margin 
of both lobes of the seed being filled up by ‘the thick 
kernel. Imported varieties of H. striatifoliwm give reason 
to helieve that it is one species with reticulatum. There 
are at present thirty-five different hybrid crosses in the 
genus [Hippeasirum at Spofforth, and four or five more 
other collections; but every attempt to obtain a mule by 
~ the pollen of any other genus applied to Hippeastrum has 
failed. The pollen of all these mules appears to be fertile, 
and three have produced seedlings by their own pollen, 
viz. regine-vittatum, fulgido-rutilum, and rutilo-regmé 
vittatum. It is smgular that Hippeastra appear to be 
more easily fertilized by the pollen of another ‘species, 
of a mule, than by their own: for instance, one flower 
subbarbatum bore seed by rutilo-vittatum, and one 
rutilo-vittatum by subbarbatum, and the other flowers 08 
both plants touched with their own pollen were sterile. 
Similar results have been ‘frequently. observed in this 
genus. W, H. 
