122 PITTONIA. 
charaeters of a rotate or campanulate corolla, and free and de- 
clined stamens were virtually abandoned for Polemonium when 
P. confertum was admitted; and the genus was then ac- 
knowledged by Dr. Gray, to have no other foundation than 
that of its peculiar habit! We think it has somewhat more 
- than that to rest on, though not so very much more. : 
The three genera to be discussed in the present paper differ 
from the others in the order, in respect to the relations of the 
fruiting calyx to the fruit itself. After flowering, the calyx 
continues to grow, more than keeping pace with the develop- 
ment of the capsule, so that when the latter is mature it is 
scarcely in contact with the former. This characteristic, Pole- 
monium, Collomia and Navarretia have in common. In Phlox 
and in Gilia the calyx becomes distended by the capsule be- 
fore its maturity, and is eventually ruptured by it. 
Polemonium is most nearly related to the Ipomopsis section 
of Gilia. This is well shown in that very natural arrange- 
ment, or ordering of the genera which Dr. Gray has given us 
in the Synoptical Flora. But in my view, Polemonium is to 
end where the calyx ceases to be herbaceous and flaccid to the 
very tips of its segments, and where the latter cease to remain 
erect, above a tube which is not costate or with hyaline inter- 
vals below the sinuses. 
Collomia, restricted to species well at agreement in habit, is 
marked, as will be seen, by a character precisely analogous to ' 
that which so good a genus as Nemophila rejoices in among 
the Hydrophyllace:e, though in Collomia the mark is absent 
at flowering time, making its appearance later on, but is always 
conspieuous in the fruiting plant. The calyx here is also 
more or less scarious below, and distinctly angular; the seg- 
ments perfectly equal and rigid, often aristate-pointed. 
-~ In Navarretia we have the ealyx of Collomia without the 
replieate sinuses and with a very irregular limb, two or three 
of the segments being two or three times larger than the other 
two or three, the smallest being aristate-pointed, the largest 
' Syn. FL Vol. ii. part 1. p. 150. 
* 
