THE GENUS CHAMACRISTA. 239 
state that in all normal and typical Cssalpineous Legumi- 
nose the flower, though almost regular and rosaceous, so 
to speak, as to its corolla, is nevertheless just like the papil- 
ionaceous flower in its plan. There is in the corolla of 
both one odd petal which is uppermost and may in either 
family be called the banner or vexillum. Next below the 
banner comes a pair of petals called the wings; then a 
lowermost pair denominated the keel. In shorter phrase, 
the ordinary cæsalpineous corolla, though quite regular, is, 
like the papilionaceous corolla, made up of banner, wing 
and keel petals; the banner being always an odd petal and 
uppermost in position. Now the puzzling circumstance in 
the case of my Chamaerista corollas, as newly observed after 
having become familiar with those of real Cassias, lay in 
the obvious fact that, in the place of one lone petal at the 
top of the flower where the banner should be, there was a 
pair of petals. Then at the base of the flower, where a 
closely matched pair should be, in place of the papilionaceous 
keel-petals, there was seen to be a solitary petal larger than 
any of the others. At thig point I seemed at first to have 
solved all difficulty by assuming that in this species the 
flower is inverted, turned up side down by a probable twist 
in its pedicel. Such inversion of corollas is not rare in the 
plant world ; though it never oceurs in the same genug with 
plants having flowers in normal posture. It is the mark 
more often of an entire family or natural order than even 
ofa genus. The Violaces, the Lobeliaceze and the Orchi- 
dace; are familiar examples of large families, one of the 
essential characters of each of which is an inverted corolla. 
Logically, therefore, this new discovery would separate the 
Chamecrista group, if not altogether from the family of 
the Cæsalpineæ, at least from Cassia, to form a distinct 
genus. Yet hitherto, as another day’s observation proved, 
I was in error; and my perplexity deepened as I was driven 
from my assumed theory that the Chameerista corolla is in- 
verted. That larger odd petal at the bottom of the flower 
32 
