18 CONTRIBUTIONS FROM THE NATIONAL HERBARIUM. 
degree of difference between this and any other species of the group 
is much greater than that between such species as P. fasciculatum, 
P. arizonicum, and P. adspersum. But other groups, such as the 
Lanuginosa, are made up of closely allied species, connected with one 
another by intergrades, to form a composite taxonomic network whose 
component parts cannot be definitely distinguished by clean-cut 
lines of demarcation. 
DESCRIPTION OF THE GENUS AND SPECIES. 
PANICUM L. 
Panicum L, Sp. Pl. 55, 1758. 
Steinchisma Raf. Bull. Bot. Seringe.220. 1830.4 
Phanopyrum (Raf.) Nash in Small, Fl. Southeast. U.S. 104. 1903.6 
Spikelets articulated below the glumes, falling entire, more or less compressed 
dorsoventrally, arranged in panicles, rarely in racemes; glumes two, herbaceous, 
nerved, usually very unequal, the first often minute, the second typically equaling 
the sterile lemma, the latter of the same texture and simulating a third glume, bearing 
in its axila membranaceous or hyaline palea and sometimes a staminate flower, the 
palea rarely wanting; fertile lemma chartaceous-indurated, typically obtuse, the nerves 
obsolete, the margins inrolled over an inclosed palea of the same texture, a lunate 
line of thinner texture at the back just above the base, the radicle protruding through 
this at germination; stamens three, styles two, stigmas plumose; grain dorsoventrally 
compressed, with a punctiform hilum, free within the firmly closed lemma and palea. 
Annual or perennial herbaceous grasses, of various habit, confined to the warmer 
regions of both hemispheres. 
A number of species here included in the genus Panicum depart in some measure 
from these generic characters. Thesubgenus Paurochaetium approaches Chaetochloa 
Scribn. and the group hitherto referred to Panicum section Ptychophyllum A. Br., in 
that the uppermost spikelet of each branchlet is subtended by a bristle-like prolonga- 
tion of the axis. Panicum geminatum Forsk. and P. paludivagum Hitche. & Chase 
have a racemose inflorescence as in Brachiaria Griseb., but the spikelets are placed 
with the back of the fruit turned toward the rachis as in true Panicum, not in the 
reverse position asin Brachiaria. In P.barbinode Trin., P. arizonicum Scribn. & Merr., 
and P. tecanum Buckl. spikelets toward the ends of the branches are often placed in 
the reverse position characteristic of Brachiaria, while others on the same branch are 
placed with the back of the fruit toward the axis, showing that in an inflorescence 
not strictly racemose this character of the position of the spikelets in relation to the 
axis is not of taxonomic significance, since it depends on whether one or the other of a 
pair of spikelets on a one-flowered branchlet has been developed. Hence this charac- 
ter, while distinguishing between Paspalum L. on the one hand and Axonopus Beauv. 
and Brachiaria Griseb. on the other, does not alone clearly separate the latter from 
Panicum, but must be taken in connection with the strictly racemose inflorescence. 
In Panicum elephantipes Nees the thin but not hyaline margins of the acuminate lemma 
are not inrolled above the middle, the fruit thus suggesting an approach to Valota 
Adans., but in texture it is not cartilaginous and papillose as in that genus nor does 
P. elephantipes approach Valota in habit or inflorescence. In the Verrucosa and the 
related Trichoidia the firm-margined lemmas are not inrolled except at the base. 
a See discussion under P. hians, p. 118. 
b See discussion under P. gymnocarpon, p. 327. 
