402 FLOWERS OF ANTHOXANTHUM—HOLM. 
_ 
It may not be unreasonable to suppose now, that the spikelet of Anthow- 
anthum has three flowering glumes, although we have been unable so” 
far to observe any trace of a palet or rhachilla i in the axil of the lowest — 
situated of these glumes. 
We now want to reply to the next question and show whether the per- 
fect flower, the uppermost one in the spikelet, is terminal, as stated by 4 
Dll, Kichler, and Schumann, or lateral. It is hardly necessary to offer 
any farther diecusGod concerning this point, since it is a simple conse- 
quence of what has been shown in the two spikelets—Figs. 6 and 8; — 
because we have seen in Fig. 8 that a palet is present, and thereby a — 
rhachilla indicated, besides that in both figures the ace is distinetly 
elongated above the base of the flowering glume (Fig. 7), and shows 
here a pointed processus, as usual in the spikelets of the Graminece. 3 
Although, as stated above, abnormally developed specimens have 
been used to demonstrate the morphological identity of the organs — 
in the normal spikelet of Anthoxanthum, it may not be denied that the 
comparison seems to favor the generally adopted idea that the spikelet 
really contains three flowers, but of which only one is perfect, and that 
this same flower is not terminal but lateral. Before leaving the sub- 
ject we wili call attention to a very peculiar malformation observed in — 
some of the examined specimens. An abnormally developed flowering | 
elune of one of the neutral flowers (Fig. 11), the apex of which showed 
not only a long and strongly bent awn, but also on each side of this awn 
was an appendage, the structure and shape of which reminds one very 
much about the glume itself. We have then in this glume the same 
kind of prolification, of which similar cases have been recorded by 
Masters.* 
The large number of truly viviparous spikelets examined did not 
show anything of particular interest; 1t might be noted that in these, 
asin all the other malformed spikelets, the empty glumes had con- 
stantly preserved their usual and normal appearance. 
U. 8S. NATIONAL MusEvUM, 
Washington, D. C., March 21, 1892. 
ee aa 
ee 
Nadie oe 
EXPLANATION OF PLATE XLVIII. 
Anthoxanthum odoratum Li. 
Fig. 1. Inflerescence with proliferous spikelets, natural size. 
Fig. 2. Proliferous spikelet from the apex of the inflorescence, magnified about four 
times. The empty glumes are normally developed, while the flowering 
glumes of the neutral flowers are reduced to one; the rhachis is strongly 
elongated and bears at its apex three spikelets, the median nearly sessile, 
the two lateral ones distinctly pediceled, all surrounded by normal empty — 
glumes. 
Fig. 3. Spikelet a of the preceding, showing a normally developed spikelet with — 
one pair of empty glumes, two awned flowering glumes of the neutral 
flowers, and a perfect flower with the flowering glume and palet of the- 
same shape. ¢.4 X natural size. 
‘ , 
ae a ee ey ee ree ee ven Feeney pene te tee a ws ae eee 
* Maxwell Masters: Vegetable Teratology, London, 1869. ; 
