392 
G. Brandegeei was described by Dr. Gray from material col- 
lected by T. S. Brandegee in northern New Mexico. The essen- 
tial features separating it from G. crifidum were supposed to be its _ 
peculiar matted cespitose habit, short stems, small fleshy leaves, 
and short mostly axillary pedicels. An examination of the type 
shows that many other Rocky Mountain and Pacific Coast speci- 
-mens must on this basis be included in this species. Specimens 
from California, Wyoming, Colorado, Nevada and the Saskatche- 
wan are practically identical with the Brandegee type. G. Bran- 
degeei seems to be a good species, although rather closely related 
to G. trifidum. Its essential features are as stated above, the ces- 
pitose habit, generally smooth stem, broad and slightly fleshy 
leaves, and short glabrous one-flowered lateral pedicels. . 
G. arcuatum is here described as new. It includes a peculiar 
group of plants somewhat related to G. dzfolium Wats., as well as to 
G. twrifidum, but it is quite distinct from any described species. 
The distinguishing features are the long rather obtusely angled 
stem, acute leaves, and short stout recurved lateral pedicels.: 
Specimens were distributed by Mr. Howell named G. trifidum var- 
cuspidulatum Gray; otherwise this species has never received a 
name. 
G. tinctorium L. has long been an ambiguous species. Many 
botanists have attempted to separate it, but so far with very little 
success. From the description given by Linnaeus one can scarcely 
tell which plant is meant, but relying upon Willdenow, it is evi- 
dent that this name must be applied to our large-leaved smooth 
eastern form with four-cleft corolla. G. “inctorium is perhaps the 
most variable of any species of this group. A careful comparison 
of a great many specimens, however, enables us to separate it into 
four well-marked and fairly distinct series, but which are too closely 
related to allow of specific distinction. The type is here taken to _ 
be the ordinary form of the Eastern States, with large leaves vary- 
ing from linear to lance-ovate. Descriptions given by the early 
authors fit this better than any other, and besides it is the one 
most likely to be met with by the early travelers. This type in- 
cludes the G. tifidum latifolium of Torrey and later authors. 
Specimens intermediate between the broad-leaved form and the 
narrower are more numerous than either extreme. A very dis- 
