[Vor. 4 
86 ANNALS OF THE MISSOURI BOTANICAL GARDEN 
something like this has happened in the origin of many 
genera. 
The differences between Trillium and Paris may, ав we 
have seen, be reduced to three; but these, so far as we know 
from present variations, are apparently independent of each 
other as arule. Three mutations are required to account for 
the origin of a typical Paris from a Trillium (two if we con- 
sider P. tetraphylla). Other intermediate species containing 
one or two of these features only have recently been described 
by Léveillé from China. Thus P. Dunniana Lévl. and P. 
aprica Lévl. also have the anther connectives scarcely, if at 
all, prolonged, while in P. atrata Lévl. the petals are longer 
than the sepals. We may, therefore, assume a considerable 
amount of elimination of such forms, perhaps through their 
own instability in inheritance, until finally a stable combina- 
tion was reached which has sinee given rise to the various 
species of Paris through another group of variations. It is 
greatly to be hoped that some one will undertake crossing 
experiments with Paris and Trillium, for they would throw 
much light on these questions. 
MEDEOLA Gronov. 
А monotypie genus of eastern North America. 
1. Medeola virginiana L. Sp. Pl. 339. 1753. 
Gyromia virginica Nutt. Gen. 1: 238. 1818; Lamarck, 
Encye. Meth. 4: 4. 1796; Illustr. Gen. Tab. 2: pl. 266, fig. 2. 
1823; Barton, Elem. of Bot. pl. 14. 1803; Curt. Bot. Mag. 
pl. 1316. 1816; Meehan, Native Flowers 2: 157. pl. 40. 1879. 
Nova Scotia and New Brunswick to Ontario, Minnesota, 
Florida, and Tennessee. 
The genus Medeola is remarkably distinct from its nearest 
relative, T'rillium, yet there is no question of its affiliation, on 
the one hand with Trillium and on the other hand with Paris. 
The differenees enumerated below would seem to indicate 
that Medeola is the sole survivor of a group of North Amer- 
iean forms which has disappeared. 
