AMARYLLIDACEA2. 
295 
horizontal position ; 1 find the like position in the vernal 
species of Narcissus, and a sloping posture also in Queltia. 
In Corbularia and Hermione I find the fruit erect from the 
time it begins to swell. There is a difference of stamen in 
the jonquil and an erect capsule, and the autumnal Narcisseae 
have also an erect capsule, and I am doubtful whether, on 
close inspection of the latter, they may be found to disagree 
with Hermione, and perhaps form a different genus with the 
jonquil. I find the seed of Ajax roundish, wrinkled with an 
elevated wrinkled raphe and chalaza; of Ganymedes ap- 
proaching to that of Ajax ; of Corbularia irregularly flattened, 
wrinkled, and somewhat foliaceous ; of Hermione shining, 
smooth, properly compressed with a rounded back, irregu- 
larly angular by contact, the margins of the angles sulcate 
on the side, and the raphe sulcate with elevated margins ; 
that of Narcissus roundish, sometimes angular by contact, 
minutely lined or wrinkled with an elevated raphe and 
chalaza. I have not however had sufficient opportunities 
of tracing the diversities of the seed, and I have seen so 
much outward difference of appearance between the seeds of 
two Hermiones, that I cannot describe the seed of Hermione 
with any confidence. I have not been able to obtain that of 
any Queltia, except the doubtful species Jonquilla, which 
agrees with the autumnal Narcisseae in having an erect cap- 
sule. I have admitted the genus Ganymedes with very great 
hesitation, having been disposed to look upon it as a section 
of Queltia. It forms, however, a very marked group, and I 
find no intermediate plants to connect them. Mr. Salisbury 
had a remarkable intuitive perception of differences which 
has generally proved to be correct, though he was unable to 
give any satisfactory account of the existing marks of dis- 
tinction. I am indebted to him for nothing in this natural 
order, but naked names of genera, excepting Acis and Car- 
polyza which he defined in the Parad. Lond. He brought 
his MS. to me one morning, and read to me some names and 
observations of little importance from it, but did not let me 
look into it, and I believe it contained no finished generic 
characters. He could give me no account of the features on 
which he built his separations, except that he had founded 
Hymenocallis on two-seeded cells, which was quite incon- 
sistent with the fact, the range of the genus extending from 
2 to 8-seeded cells. The names of the Narcissean genera and 
Eurycles were published by him in the Hortic. Trans, with- 
