1904] CHRYSLER—CENTRAL CYLINDER 77 
Viewing in a comparative way the genera of Liliaceae described 
in the foregoing paragraphs, it appears that Trillium exhibits very 
clearly the stages in development of the stele. These stages may be 
briefly enumerated as follows: (1) the protostelic condition is present 
in the basal part of the stem and persists through one or more inter- 
nodes; then follows (2) the siphonostelic condition in which cortical 
tissue is included in the stele above the point of exit of the leaf traces 
and thenceforth forms a medulla; (3) many segments of the stele take 
on the amphivasal character; (4) strands of vascular tissue, usually 
amphivasal, turn into the medulla where they run for a greater or less 
distance and may become connected with leaf traces. Though the 
stem of Trillium seldom shows any traces of a phloeoterma, Clintonia 
borealis presents a diagrammatic example of a stele which never gets 
beyond stage (3), and has external and internal phloeoterma which 
communicate through the foliar gaps. The internal phloeoterma is 
probably degenerate in Maianthenum except at the edges of the leaf 
gaps of the young stele; there may be a physiological correlation 
between the very heavily cutinized external layer and the absence of 
an internal layer; stage (4) is much delayed in this plant. In Smila- 
cina stellata stages (3) and (4) appear sooner; the phloeoterma is less 
distinct. Medeola and Lilium show the effect of long internodes 
combined with extended gaps in breaking up the central cylinder 
Into several strands arranged on the circumference of a circle. Uyu- 
laria and Streptopus quickly pass into stage (3). Many members of 
the family such as Allium have assumed the bulbous habit, and in 
the very short stem of these plants the medullary strands appear very 
early. They probably express the highest order of specialization 
shown in the family. 
As to the bearing of the foregoing observations on the central 
cylinder of the two families upon so-called stelar theories, it may at 
Once be stated that though the pteridophytes must be the critical 
era in any discussion of these theories, yet information from even 
Ye highly Specialized a group as the monocotyledons is of importance, 
if we acknowledge the descent of the seed-plants from fern-like ances- 
‘ors. Many of the Liliaceae studied do not seem to afford any evi- 
dence on the points in dispute, which is not to be wondered at when 
