1898] BRIEFER ARTICLES 201 
than the anterior and are also probably inserted somewhat higher on 
the axis. Consequently the lodicules, when all are present, continue 
the distichous arrangement of the palet and glume.” 
In accordance with these conclusions, Hackel formulated his inter- 
pretation of the organs of the flower in grasses. It may be noted that 
Hackel expresses doubt about his interpretation of the significance of 
the posterior lodicule. 
Usually the bamboos have the posterior lodicule present and, in a 
lew species, there normally develop two trimerous whorls of lodicules. 
Last summer a bamboo (Arundinaria falcata) produced flowers in the 
University garden and the writer took the opportunity to study the 
flowers in detail with special reference to Hackel’s conclusions. 
Warming in his Systematic Botany (p. 291) regards the Bambusez 
asthe most primitive of the tribes of the Graminez, and it seems to 
me that a fair consideration of their floral structure and their geograph- 
ial distribution bears out his opinion. He also states the theory of 
the grass flower substantially as the writer is about to present it, but he 
gives this theory in brackets after having in the preceding paragraph 
given the bract theory in italics. If the bamboos be the most primi- 
ie of the grasses, then surely the lodicule in this group of plants 
oat to shed light on the question of their morphological signifi- 
e. 
: pis 
Serial 
is Pid and longitudinal sections of the flower were made and 
This ? are representations of what was found. 
Sccur in the ovary walls the three midribs of the coherent 
