1904] CHRYSLER—CENTRAL CYLINDER 169 
development of this corm has been accompanied by considerable 
changes in the vascular system, leading to complications which 
render this plant unsuitable for the purposes of the present inquiry. 
ARISAEMA DRACONTIUM, A, SPECIOSUM, A. INTERMEDIUM, and A. 
TARTARINOWII all resemble A. ériphyllum in having seedlings which 
show a complex network of bundles. They all likewise produce a 
corm. 
TYPHONIUM DIVARICATUM has a vascular system so similar to that 
found in Arisaema that it does not merit a separate description. 
In viewing in a general way the genera so far described the ques- 
tion arises: What characters are to be regarded as primitive? The 
answer must be, those which occur in the first formed part of the stem, 
unless there is reason to believe that this region has been influenced 
by the assumption of some special habit, such as the tuberous or 
bulbous habit. The stem of Acorus is relatively free from external 
influences, on account of its geophilous habit; its central cylinder is 
at first protostelic, then siphonostelic with a pith communicating 
with the pericycle through the foliar gaps. Judging from the condi- 
tions in Symplocarpus and in the mature organs of Acorus, we may 
infer that if the central cylinder of the Acorus seedling were not so 
narrow the endodermis and cortex might here also enter through the 
gaps, in which case the stele would differ from that characteristic of 
the ferns mainly in the absence of internal phloem, a feature which 
appears to be quite rare in seed plants. The simple siphonostelic 
Stage persists in Acorus for several internodes, and the stem looks 
much like that of a dicotyledon; higher up some segments of the stele 
become amphivasal, and this may be regarded as the first appearance 
of a monocotyledonous character; very soon certain strands begin — 
‘o run in the medulla, and so the monocotyledonous nature of the 
stele is established. The steles of the various genera differ from the 
type Just described in a modification of the basal part of the stele in 
accordance with the tuberous habit, as in Symplocarpus, or in the 
tapid disappearance of the phloeoterma, as in Peltandra, or in the 
early appearance of the medullary strands, as inj Arisaema. What- 
“ver may be the nature of the pith in Acorus, there seems to be good 
reason for believing that in Peltandra and Symplocarpus the pith is 
