140 CONTRIBUTIONS FROM THE NATIONAL HERBARIUM. 
Panicum lanatum Sw. Prod. 24. 1788. , 
The type specimen marked °P. lanatum fl. ind. oce. Jamaica. Swartz,’’ is allied 
to P. divaricatum L. and P. sloanet Griseb. The species is characterized by the 
densely lanose-velvety sheaths and blades. On account of the earlier Pantewm 
lanatum Rotth. Deser, PL. 3.1776 ( Valota insularis (L.) Chase), I suggest for Swartz’s 
species the name Panicum swartzianum (Panicum lanatwm Sw., not Rottb.). 
Panicum arundinaceum Sw. Prod. 24. 1788. 
There are two plants upon the type sheet, which is marked ‘ Panicum arundina- 
one with a large spreading panicle, the other with 
’ 
ceum fl. ind. occ, Jamaica. Swartz,’ 
a narrow compact panicle. The description applies better to the first, although they 
are both Jsachne arundinacea (Sw.) Griseb. 
Panicum polygamum Sw, Prod. 24. 1788. 
The type specimen marked * Prodr.”’ is Panicum maximum Jacq., an earlier name, 
which Swartz himself uses in his Flora.¢ 
Panicum glutinosum Sw. Prod. 24. 1788. 
The type specimen marked °° 2. glutinosum fl. ind. oce.”? from “Jamaica. Swartz,”’ 
belongs to this species as generally understood. 
Panicum trichoides Sw. Prod. 24. 1788. 
There are two plants upon the type sheet, which is labeled “ trichoides fl, ind. oce.”’ 
from ‘Jamaica. Swartz.’’ The left-hand specimen is the form described as P. brevi- 
folium in Grisebach’s Flora; the other is the same as the type of Panicum tricanthium Nees 
in the Berlin Herbarium,examined at Halle. The leaf blades of the right-hand speci- 
men are longer and the spikelets somewhat larger than in the left-hand one. Sloane's 
plate? of Panicum brevifolium is cited by Swartz. Itis evident from the more complete 
description given later in his Flora that Swartz considered his species the same as /’. 
brevifolium L. and variable enough to include both the plants preserved, but wished 
to change the name. The type of P. brevifolium L. is from India. The species is 
shown by description and the specimen in the Linnzean herbarium to be P. ovalifolinm 
as described in Hooker’s Flora of British India. Hence we may consider P. trichoides 
Sw. as applying to the Tropical American species usually described as P. brevifolium 
L. (P. ecapillaceum Lam, Tabl. Encyel. 1: 173. 1791), the left-hand plant being taken 
as the type, and adopt Panicum trichanthum Nees for the larger form. 
Panicum caespitosum Sw. I*l. Ind. Occ. 1: 140. 1797. 
The type specimen marked “ 7’, caespitosum fl, ind, occ.’ from “Jamaica. Swartz” 
is Panicum prostratum Lam., but, as shown previously (page 119), we should use the 
name Panicum reptans L. for this species. The Mexican plant which has been dis- 
tributed under the name 2. caespitosum in recent collections is a different species. ¢ 
@Fl, Ind. Oce. 1: 170. 1797. 
bHist. Jam. pl. 72. f. 8. 
¢ This was brought to my attention by Professor Mez, of Halle, for whom I propose 
to name this species: 
BRACHIARIA MEZIANA sp. nov. 
Perennial; culms cespitose, glabrous, at first erect, 20 to 30 em, high; later branched 
and decumbent becoming as much as 70 cm. long; leaves light green, sheaths densely 
ciliate on the margin, sometimes sparsely pilose on the surface, blades 5 to 12 em. long, 
5 to 10 mm. wide, moderately stiff and firm, glabrous or sparsely pilose on either sur- 
face, ciliate on the margin near the base with papillose hairs; early panicles long- 
exserted, later ones less so or scarcely exserted, consisting of several spike-like racemes 
2 to 3 cm. long, along the upper 2 to 3 em, of the culm; spikelets placed with the 
first glume toward the axis, arranged in 2 rows on one side of a somewhat flattened 
narrow rachis interspersed with pilose hairs, nearly sessile, glabrous, 3 mm. long, 
ovate, subacute; first glume ovate, 3-nerved, | mm. long, second glume as long as 
