10 BOTANICAL GAZETTE [JANUARY 
may continue through several internodes. This probably accounts 
for the prevalence of these bundles in the rhizomes of monocotyle- 
dons, where they were first observed. A comparison of the aerial 
and subterranean nodes of the grasses under study has not yielded 4 
results of great significance; most species show no noticeable differ- 
ence in the number of amphivasal strands in the two cases. But 
in Andropogon jfurcatus, A. scoparius, Chrysopogon avenaceus, 
Zizania aquatica, and Phleum pratense, the amphivasal strands are 
distinctly more numerous in the basal nodes. No examples of the 
opposite condition have been found. Querva found in Gloriosa 
(7) that the amphivasal bundles are connected with the origin of 
a branch. Among the grasses, Phalaris arundinacea, Paspalum — 
stoloniferum, Sporobolus Wrightii, Coix lachryma, and Zea Mays : 
show these bundles at the point of origin of branches, but in other 
species only collateral bundles could be discovered at these places. 
Too much importance should not be attached to the few cases named, 
in view of the fact that the sedges uniformly show amphivasal bundles 
associated with leaves and not with branches, as PLOWMAN has _ 
shown (10). This is one reason for considering the grasses a more 
specialized group than the sedges; in fact it may be premised that 
the amphivasal condition originally connected with leaf-traces has 
in the Gramineae spread to the branches. The occurrence of amphi- 
vasal bundles in the leafy reproductive axis of Zea, while they are 
rare in the main stem, deserves emphasis. While many of the 
grasses show amphivasal bundles in all the nodes, in this highly | 
developed genus the bundles in question have nearly disappeared 
from the ordinary nodes, but .have persisted in the conservative 
region named. STRASBURGER proposes (4, p. 348) a physiological 
explanation for the occurrence of these bundles, viz., that the amphi- 
vasal structure is favorable for the taking up of reserve materials 
stored in a rhizome, but this explanation is not in accord with the 
accepted view that elaborated food is carried not by the xylem, but 
by the phloem. Examination of serial sections leads to the opinion 
that the mechanical necessities of bundle fusion rather than con- 
siderations of absorption of food have been the determining factor 
in producing these bundles. 
If we accept the view advanced above, that these bundles are to 
