158 
specting the rank to be given to these three well-marked groups 
of species is not a matter of authority, nor of personal feeling or 
sentiment, but one simply of good judgment based upon sufficient 
knowledge. As to my own opinion, after eleven years during 
which the subject has been several times under consideration, I 
still think that the separation of the genus Megarrhiza is the 
most satisfactory course to take. I also think that Naudin’s 
genus Echinopepon should be re-established. The three genera 
appear to me to be sufficiently well distinguished by the following 
characters :— 
ECHINOCYsTIS. Annual; germination epigzous, as usually 
in the order; flowers hexamerous; fruit ovoid, bursting irregu- 
larly at the apex, two-celled and the cells two-seeded ; seeds flat 
and thin, rugose, oblong with an attenuate base.—One species, 
of the northern Atlantic States and Canada. 
MEGARRHIZA. Perennial, with a tuberous root; germination 
hypogzous in a peculiar way (so far as known); flowers normally 
pentamerous; fruit ovoid or globose, bursting irregularly at the 
apex, two-celled or four-locellate, the cells one-four-seeded ; seeds 
turgid, rounded at both ends, smooth.—The five species of the 
Botany of California, with 17. macrocarpa, Rattan, and M. Gilensis, 
Greene; the last species in Arizona. 
ECHINOPEPON. Annual; germination epigzous; flowers 
normally pentamerous; fruit oblong, attenuate at both ends, 
dehiscing at the apex by a deciduous operculum, two-celled, the 
cells four to six seeded ; seeds small, flattened, rugose.—&. horri- 
dus, Naud., and E. Wrightii, of New Mexico, Arizona and north- 
ern Mexico, with several other species of still more southern habitat. 
There still remain three peculiar and doubtful southern species, 
which have been referred to Echinocystis, and which may as 
well remain there provisionally until more is known respecting 
them, viz.: E. (?) Bigelovii, Cogn. (Elaterium ? Bigelovii, Wat- 
son), £. minima, Cogn. (Elaterium ? minimum, Watson, first 
described by Kellogg, in 1859, as Marah minima), and E. (?) 
parviflora, Watson. Of the three genera these species should be 
placed in Echinopepon, but it is probable that they will prove 
referable to some other genus, 
