166 
ROYAL HORTICULTUBAL SOCIETY. 
flowers of L. davuricum, contrasted with the absence of hairs, 
the lanceolate pointed leaves, and the externally glabrous flow^ers 
of L. hulbiferum. These characters, however, are often insufficient 
to diagnose the species, because in gardens so many interme- 
diate forms and varieties present themselves (possibly originated 
through hybridization) that one may often be in doubt as to 
the species in the absence of bulb or capsule to settle the 
matter*. 
As I believed I had noticed that in L. hulbiferum the scales 
of different bulbs sometimes exhibited a very indistinct contrac- 
tion above the base, while sometimes it was distinct, it occurred 
to me whether in the last case I might not have had a hybrid 
between this species and L. davuricum. 
I fertilized therefore a flower of L. hulbiferum with the pollen 
of L. davuricum, and a flower of this with the pollen of the first. 
This was the only fertilization eff'ected, because these individuals 
bore only a single flower each, and I neglected the operation 
in the case of others. The ovaries of both species swelled up 
and developed. But after the capsule of L, davuricum had be- 
come about two inches long and 5 lines in diameter, it ceased 
further growth and dried up. It had, however, sufficiently 
developed itself to show the characteristic form of the capsule 
. of L. hulbiferum, the male parent. On the other hand the capsule 
of L. hulbiferum grew and ripened, and was not cut ofl" till it had 
dehisced ; it had completely the character of the male parent-^ 
that is, of L. davuricum. 
Both species, then, had so completely changed their capsules, 
that, having forgotten to which the long and deeply umbilicate, 
and the short and flat one belonged, I did not observe the won- 
derful state of the case till I was about to incorporate the corpus 
delicti into the herbarium, and compared it previously with descrip- 
tions and other fruit-specimens. This inattention of mine was 
very blamable, because, believing the other capsule which did not 
complete its growth of no value, I allowed it to drop. The 
developed capsule I have carefully preserved : it contained about 
130 seeds ; a few, not exceeding a dozen, may have been lost. 
The breadth of the wing was about the mean of what it is in the 
* ]^Baker, in fact, describes L. hulbiferum proper as having the stem, especiallj 
in the upper part, clothed with scattered pubescence, and the perianth shghtlj 
cottony externally, while in L. davuricum the stem is nearly smooth. — Tr.] 
