206 Rhodora [DECEMBER 
England Botanical Club was re-examined and quickly found to be 
similarly divisible. The form with purple-tipped scales proved to 
have a generally coastal distribution from eastern Massachusetts to 
Florida, Central America, Colombia, Guiana, and the West Indies, 
with a single outlying specimen from New South Wales, while the form 
with attenuate white-tipped scales extends across the North American 
continent, showing some variations to the westward, and being also 
frequent as an introduced plant in various parts of the Old World. 
Study of the literature of the group leaves no doubt that the widely 
distributed plant with white-tipped scales is the typical E. canadensis 
L., while the coastal plant with purple-tipped scales is the E. pusillus 
of Nuttall, who long ago recognized its lower stature, smoother stem, 
and more slender, open, and fewer-headed inflorescence, as well as its 
consistently entire leaves. "The small purple dot at the apex of the 
involucral scale, though a valuable differential character, is incon- 
spicuous except under a lens, and seems to have been overlooked by 
Nuttall, but all characters mentioned by Nuttall correspond perfectly. 
with the plant in question. Furthermore, there is confirmation in 
the identity of a specimen labelled E. pusillum Nutt., now in the Gray 
Herbarium, which was received from the herbarium of Nuttall ap- 
parently at the time when Torrey and Gray were preparing the manu- 
script of the Compositae for their Flora of North America. It is 
interesting to find that this doubtless authentic specimen exhibits not 
only all the characteristics attributed to his species by Nuttall but 
shows the purple-tipped involucral scales. 
The examination of many specimens from numerous widely separ- 
ated stations shows the same distinctions between E. canadensis and 
E. pusillus, and as yet no tendency toward intergradation has been 
observed. It seems therefore that Nuttall's E. pusillus may be re- 
instated as a valid species, near to, though readily distinguishable 
from, E. canadensis. It will be interesting to learn whether its range 
can be extended further toward the northeast. 
The only attempt, which I have found in recent literature, to re- 
instate E. pusillus Nutt. is apparently a mistaken one. It is the Lep- 
tilon canadense pusillum (Nutt.) Daniels, Flora óf Boulder, Colorado, 
239 (1911). No description is given beyond the statement: “The 
common form of the foothills, 3-1 dm. high and but few-flowered, 
6000-8000 ft. (Daniels, 694)." An examination of a considerable 
number of specimens of E. canadensis L. from the Rocky Mountain 
