164 BOTANICAL GAZETTE [FEBRUARY 
ately outside of the vascular system is pith. The xylem and 
phloem are therefore, for the most part, placed in the usual collateral 
arrangement, with phloem next to the cortex. The protoxylem 
occurs next to the pith. Hence we may describe the vascular 
system of the corm as an everted ectophloic siphonostele with 
endarch xylem. The vascular cylinder therefore differs from that 
of the rhizome only in being everted. 
The constant occurrence of a double leaf trace in this specimen 
is noteworthy. Exactly the same situation has been described for 
O. palmatum by Bower (3). The same writer (2) describes the 
leaf trace of O. pendulum as composed of five strands, and thinks 
it probable that this is the case in all the species of the group 
OPHIODERMA. If this be true, the specimen described in this paper 
represents the fifth species of the genus in which a multiple leaf 
trace occurs. 
BOWER points out the fact that those species which have a 
multiple leaf trace are offshoots of the main line of the genus, and 
he considers them as derived forms. In the case under discussion, 
the specimen described belongs to the species usually considered 
most representative of the genus. It is to be noted that the 
rhizome of O. palmatum is very broad in proportion to its length, 
and often almost spherical in shape. While the rhizome of O. 
pendulum is relatively small, the figures of O. simplex indicate a 
rather stout stem, and CAMPBELL (8) describes the rhizome of 0. 
intermedium as “very short, forming a small tuberous body.” It 
seems probable that in the protocorm described above the doubling 
of the leaf trace is related directly to the extreme lateral expansion 
of the corm, due to its manner of growth; and it is possible that in 
the other species of this genus with a multiple leaf trace the same 
explanation may hold. 
If this explanation of the doubled leaf trace is accepted, every 
unusual feature of the protocorm is definitely related to the peculiar 
behavior of the segments of the apical cell. It is to be noted that 
these segments in the protocorm differ in their development from 
those in the rhizome in two particulars: (x) in the rhizome the 
first division of a segment is transverse, and the later divisions are 
variable; in the protocorm the first two divisions are longitudinal; 
