1895.] North American Amaranthacee. 341 
tending southward into Mexico (Hemsley Biol. Centr. Am.) 
and from the West Indies (Mog. |. c.). The only specimen 
in our possession found west of El Paso, Tex., was collected 
by Pringle near Chihuahua, Mex. While it may occur in the 
intervening regions, it is probably quite rare west of Texas. 
ype unknown. 
A 2 
G. LANUGINOSUS Sheldoni, n. var. 
Plant more robust throughout: cauline leaves larger: flowers 
rigid, flat on the ventral side: bracts short orbicular: sepals 
destitute of chlorophyll, not scarious margined.—Collected in 
July, t891, near Cash Creek, Indian Territory, by C. S 
Sheldon (no. 170). Types in Nat. herb. and herb. Gray. 
In merging G. rigidiflorus with G. tenuiflorus we have 
simply confirmed the suggestions of Dr. Watson (Proc. Am. 
Acad. 18: 144) and of Dr. Torrey (Bot. Bound. 180). Fur- 
thermore, there is no doubt that Moquin’s G. lanuginosus 
must also be included here. Hooker f. in Benth. & Hook. 
Gen. Pl. 3: 37, 39 referred Gossypianthus lanuginosus Mog. 
to Guilleminia, briefly characterizing it by the narrower lobes 
of its perianth. Even if the transfer were correct, this dif- 
ference would scarcely be counted of specific importance in a 
group of so keen susceptibility to variation as is everywhere 
prevalent in the Ammaranthacee. But a critical study of 
Moquin’s description of Gossypianthus lanuginosus discloses 
the fact that it is quite distinct from Guzlleminia densa Mog. 
in the presence of a rosette of radical leaves and of distinctly 
3-nerved acute sepals, both of which are characteristic of 
Gossypianthus. It is not probable, moreover, that so critical 
an observer as Moquin-Tandon should have committed the 
error of confusing two genera so distinct in floral character. 
Still more conclusive is the fact that no Guz/leminia with 
Not appear in our herbaria, the only reasonable explanation 
is that the facts of range were taken from the original 
Paronychia lanuginosa. Guilleminia lanuginosa Hook. f., 
then, so far as it pertains to our own boundary region is ficti- 
tious; and the plants in question that have been referred to 
under the name must be looked for in the herbaria among the 
material labelled Gossypianthus rigidiflorus and G. tenuiflorus. 
