96 BOTANICAL GAZETTE. [ April, 
In verde sided the order, p the mother. — oe to acre ure 4 
this modificat n of terms by bee ing too clos he cha racters 
Gray’s Manu fhe same may be sa adi in ainsi wd u 
In describing "Panicim capillare he da * sterile flower neutral oni ae a sae 
glume.” A glume in no way constitutes a “flower,” and in order onform 
with bois iolnenaletut generally opted, the expression ihoul d - “ third 
glume empty,” 7. e., devoid of a palea or any of the ntial organs. In the 
same ‘teseription he calls the 2d fh arate the * upper ont gent the 4th or flower- 
ing glume he ca - a “somewhat obtuse perfect flower.” In describing Stipa spartea 
~ says the glumes, meanit ig the two empty ones, are longer than the 
of these “ sales ’ is called a “flowering glume” in the characterizing the 
“lower p 
xp ug s 
le in many cases and sufficiently exact, may lead s baveribteies or confusion: 
if too generally used. It is understood to refer to the two lowest or the first 
and second glumes of the spikelet when these alone are empty glumes. When 
the structure of the spikelet will permit its use in this sense go may be no 
objection in using it, but in other cases it would be better to designate the 
g Ss umber, as Ist, 2d, 3d, et n Reimaria, Leersia, Zizania, etc., 
there are only t u in the spikelet, and, of course, both “ outer” 
glumes, in a literal sense, but one of t a fl g glum n —_ oa 
it would hardly be ot to say that there was but one “outer” glum 
the other hand in Pani and a number of other genera, there are often 7 thes 
empty glumes below the flowering oak: Here the use of the expression in question 
would le 4 to needless complicatio: 
n the synopsis of ge snera, p page 398, there are a few errors of fact or misuse 
of terms ‘that call for a revision. Beckmannia is said t o have three empty 
Meets and a ci wher. glume. fn the apa oe plant ed are four glumes, 
the two lower o mpty while the Selig and fourth are flower-bearing, or the 
a —" enclose 0 aay palea. In merican ‘pleas, so far as it has been. 
amin y three — _prath beds discovered—the two lower ones empty, 
the third fn a perfect 
enus Panicum i is patie a as having only three glumes, two empty 
e. 
and in “ fertile.” ow, the most important character in distinguishing Pan- 
icum f palum is the fact that in the spikelets of the fo 
fi es while the latter has but three. same error occurs in describ- 
ing Setaria, whi e number and character of its 
i e here also the expression “ the fl g glume with its. 
palets.” This ave been an oversight in correcting the proof: ow- 
ering — in the Andropogonee is said to be “ often bearded.” It is often 
never bearded. Th in the genus Phalaris is statea to “consist 
ot et ed palets.” ere are no true pale halaris, 
nd it is hardly necessary to say that stamens and pistil constitute t 
flo and t lumes are aged bracts subtend i Fr. there 
a r more often six glumes; two large, complicate — ones, two smaa 
ner ones enclosing the flower and tw o (sometimes only one) i 
gs are — Pm mal — or are aiaea to simple bristles. 
ynopsis, Agr and Cinna are se arated from Deyeuxia by hav- 
ing “a ns bristle aang ts praionlte te the valet Cinna pendula, the only spe- 
ies dese bro top achilla is very often rolonged behind the palea into a 
sent naked bristle, and in Agrostis humilis Naga. a well marked species found 
in Co! orado, Peete etc., there is a similar extension of the rhachilla. 
No one would look for Graphephorum flexuosum Thurb. in Graphephor 
as chavactatent wa on page 402, where it says “outer glumes nearly equaling the 
rather remote flowers.” In G. flexuosum the outer glumes are one-half shorter 
