VOL. 111.] Notes on Californian Plants. 353 
ably fix the species; indeed, they differed only in having the mark- 
ings less distinct and well defined. But on the precipitous sides of 
the surrounding grass-clad mountains, where every recess or gentler 
acclivity was a thickly set bed of these flowers, all the color character 
of the species vanished. Repeatedly I found it easy to gather from 
one of these parterres a dozen flowers, each abundantly distinct for 
a florist’s variety, and some of which, if considered by themselves, a 
botanist might well regard _as distinct species. But with all the in- 
termediate variations so profusely present the most diverse extremes 
were traceable to the original form. 
A little study resolved these many-hued varieties into two series, 
the one of lilacs and the other of yellows. In the former the range 
of color proceeded from white through varying shades of lilac to a 
deep purple, the extreme being the var. purpurascens Watson. The 
other series passed through similar gradations from very light yellow 
to a bright lemon color, which may be identified with the var. céé- 
rinus Baker.* Sometimes the petals were of uniform color through- 
out, or were shaded from light to darker tints; in others-a lilac petal 
passed into a yellow border above, or the reverse; others again were 
yellow striate with lilac, or lilac with yellow. The upper and lower 
spots of the normal flower were occasionally indicated in these va- 
rieties, but in most instances were entirely obsolete. 
The glands were uniformly densely hairy, but varied much in size 
and shape, being oblong. oval, or transversely flattened. In some 
instances they were obsolescent. All .the plants were slenderer 
stemmed and fewer flowered than in the normal form, single flowered 
specimens being common, and few exceeded a foot in height. The 
flowers varied much in size, but were generally smaller than in the 
type. The whole exhibition appeared to be an example of pure nat- 
ural variation uninfluenced by hybridization, since no other species 
was observed in the neighborhood. 
Considering how completely all distinctions of color and markings, 
or of size and shape of gland, here break down, it becomes a question 
as to what value can be placed on these characters in a group of 
closely allied species which inhabit the same geographical region. 
ps NER ee CL ar I Oe as aOR 
7 Na SS ee aap pvc 
- *{n Dr. Watson’s Revision of the N. A. Liliacea, and in the Bot. of California this is referred as a 
variety to C. duteus, Douzgl., the most natural disposition to make from herbarium specimens. But, 
considering the associations with which it grows on these hillsides it is ev.dent that Baker was right, 
if the two species are to be kept up. fs 
D { 
