COCKERELL: NortTH AMERICAN SPECIES OF HymEnoxys 471 
only field-study could determine this. Specimens collected at the 
same locality in Wyoming in different years show differences in the 
bracts, apparently due to differing meteorological conditions, most 
probably conditions of drought and moisture. The same causes 
may have something to do with the differences between specimens 
of P. floribunda collected in different years near Santa Fé, N. M. 
Prof. K. Pearson, in his recent experiments, has found it hard to 
separate the effects of different seasons from those dependent on 
genuine racial differentiation, except after studies extending over 
considerable periods of time.* 
In the Rocky Mountains there often exist in close geographical 
proximity very different soils, differing in the amount of moisture 
they contain, in their texture, in their chemical constituents, or in 
all of these things. It remains to be shown how far these soil dif- 
ferences are the immediate and direct causes of differences in plants, 
and this can only be done by experiment. Experience suggests 
that in the main the apparently different plants of, ¢. ¢., rocky 
hillsides and alkaline flats, are really specifically different in the 
majority of instances, and could not be artificially interchanged. 
This conclusion is supported by the fact that these plants are often 
so very different (¢. g., generically so) that the idea of their charac- 
ters being non-hereditary does not suggest itself, and the fact that 
they do not spread to the contrasted environment is evident. I 
should, therefore, feel very well disposed to consider P. macrantha 
and P. ligulaeflora perfectly distinct species, were it not that I 
cannot exactly state with the herbarium material before me, where 
one leaves off and the other begins. Professor Aven Nelson, who 
knows them in the field, believed them to be distinct. 
Hymenoxys Ricnarpsont (Hooker) typical 
The type of this species and of Hooker's genus Picradenia 
came from the most northern locality (about 53° N. Lat.) in which 
Hymenoxys has yet been found. In two particulars the figures in- 
dicate a plant differing from anything I have seen, (1) the inner 
bracts are not fimbriate, (2) the pales of the pappus are remarka- 
bly short, not half the length of the disc-corolla. I should be in- 
* Biometrika, 2: 145-165. 1903. In Ficaria Ficaria, *‘ the influence of environ- 
- »? 
ment and of season are of supreme importance and screen differences due to local race. 
