COCKERELL: NORTH AMERICAN SPECIES OF HyMENoxys 469 
superne latius, pilosum praecipue apicem versus. Pappus palea- 
ceus, paleis 6-7, erectis, ovatis, acuminatissimis, membranaceis. 
“Has. About Carlton House, on the Saskatchawan. Dr. 
Richardson. — Allied to Hymenoxys of Cassini, in the nature of the 
pappus : but very different in the involucre and habit.” 
The plate shows three rather short, branched stems arising 
from a thick and evidently perennial root ; the basal leaves numer- 
ous, with linear divisions ; the rather long naked peduncles extend- 
ing beyond the leaves, the heads large for the size of the plant, 
and not very numerous; the rays broad and conspicuous, four- 
nerved, and tridentate apically ; the outer bracts united nearly to 
the middle, the inner similar to the outer, and not or hardly sur- 
passing them; the receptacle obtuse-conical, somewhat broader 
than high; the achene stout and rather short, hairy ; the pappus- 
scales long-acuminate, but not half the length of the disc-corolla. 
: This is a plant of the type of Picradenia macrantha, * P. ligulae- 
flora” and “ P. pumila.’ Comparing it with the type specimens of 
these three, we find characters as indicated in the table on page 470. 
The roots are practically the same in all four. The foliage 
does not differ in any important respects, but is more wiry, and 
pallid, in P. pumila. According to the figure, the foliage of /. 
Richardsoni is also more wiry than that of P. macrantha or P. 
Aigulaefora. The pappus-scales in P. “igulacflora are long-awned, 
decidedly more produced than in the others. 
According to the data on the type sheets, it seems that 7. 
macrantha flowers earlier than P. ligulaeflora. P. pumila occurs 
on ‘dry clayey hills” in western Wyoming. P. 4gulacflora on 
dry clayey alkaline ridges or flats on the open plains. /. macrantha 
on open stony slopes. The two extremes of this series are P. 
macrantha and P. pumila; P. ligulaeflora is fairly intermediate, 
while P. Richardsoni leans towards P. macrantha. 
The above remarks are based only on the types of the four 
species named. When one comes to examine specimens from 
numerous localities, it appears impossible to maintain them as 
separate species, owing to the existence of various intermediates. 
This seems strange, considering how totally different P. macrantha 
and P. pumila are, and I am not certain that detailed field study 
might not modify the conclusions here reached; yet with the 
