378 
ON CROSSES AND 
five fertile stamens, and with four fertile stamens and one 
abortive ; and, as in Nicotiana the fifth stamen is always of 
a different power and inserted differently from the other four, 
I was inclined to believe them to be one genus, and I made 
a great many attempts to cross them, but in vain. On further 
examination of those plants I find a diversity of the seeds, 
which are very small, and must be examined with a mag- 
nifier. Those of Salpiglossis are angular ; see pi. 43. f. 50. 
In other respects, although its aspect is easily recognized, it 
is not so easily separated by any decided distinction from 
Nicotiana; and it has been incorrectly characterized by one 
of the stamina barren, since the fifth is sometimes entirely 
wanting, and sometimes fertile. The character of Nicotiana 
in the Bot. Mag. is quite untrue ; founded on one species, 
and inapplicable to others. That genus affords an instance 
of the unimportance of the more or less continued adhesion 
of decurrent filaments, the adhesion varying greatly in the 
different species. See the seeds of Nicotiana, pi. 43. f. 51 
and 52. They are not angular, but more oblong and reni- 
form than those of Petunia, which are nearly round, and 
pitted all over. See Pet. linearis, pi. 43. f. 48. I never 
doubted the diversity of Nierenbergia, but it stands yet un- 
defined. Dr. Lindley characterised it by the reflex uneate 
lobes of the stigma, but the stigma will prove often a trea- 
cherous feature in botanical characters ; and Professor Don, 
having found in N. aristata a stigma of which the lobes are 
very little more reflex than in Petunia, at once assumed that 
they were one genus. They are, I doubt not, substantially 
distinct, but the generic character of Nierenbergia must be, 
Tube slender cylindrical , limb wide-funnel-shaped, stamina ad- 
hesive to the style and stiyma. I have never seen more than 
one seed perfected in a pod of Nierenbergia, the rest of the 
ovules proving abortive, and the capsule so small that it is 
often overlooked when ripe ; the seed of filifolia is oval with 
the back rounded. See pi. 43. f. 47. I see nothing in 
N. aristata to reconcile it to Petunia. The lovely N. caly- 
cina has a much greater affinity to Petunia from its trailing 
inflorescence and general aspect ; but, if Petunia can be 
shewn to be one with Nierenbergia, there will remain no- 
thing to separate them from Nicotiana. I am, however, 
now almost satisfied of the diversity of Petunia and Nicotiana, 
wishing, however, cultivators to persevere awhile in the at- 
tempt to cross them ; for Petunia and Nicotiana have much 
