38 ERYTHEA. 
probably pale. Its home will be found among the mountains 
back of Santa Fé, a region examined by Fendler a half-cen- 
tury ago, and totally neglected since his day. 
* * Plants, at least when mature, with a distinct horizontal 
or ascending, usually bulbilliferous rhizome. 
D. viviparum. D. crenatum, Greene, Pitt. ii. 74 (1890), 
not Raf. Atl. Journ. 180 (1833). The earlier D. crenatum, 
of which I was quite unaware at the time of making the 
homonym, is a large species of the prairies of Illinois and 
Indiana, which needs to be investigated anew, if perchance it 
is not yet extinct. 
The type specimens of the present species were in a very 
mature state, and, about the flowers of this crenate-leaved 
plant I still know nothing. Its habitat on Mt. Tacoma is not 
in the wet meadows that lie concealed amid the deep forests 
of the lower parts of the mountain; though I dare not say 
that the wet meadow Dodecatheon may not be specifically 
identical with this which I found at the limit of trees. But 
my plant grew neither in very wet nor at all grassy ground. 
It was in the immediate shade of coniferous trees, and ina 
light soil of leaf mould. I now refer to it, with some hesi- 
tancy, specimens from several collectors which show neither 
the distinctly crenate leaves, nor the large bulblets of my 
specimens; even adding some in which the rhizome is not 
well developed; for as the plants are long-lived, and the 
rhizome is of slow growth, it is but reasonable to assume 
that plants of only two or three years, though robust, and 
flowering freely, will not show anything more than a short 
and rudimentary stage of the eventual quite distinct and 
elongated rootstock. And, as even the bulblets in the bul- 
billiferous species are not developed until the plant is past 
flowering, it were unreasonable to throw out of D. viviparum 
specimens wanting bulblets which have not passed beyond 
the flowering season. I now find one character which I did 
not at first detect in the original specimens, by means of 
which I think the species may be identified even when the 
