tgtt] SINNOTT—FOLIAR BUNDLE 269 
cycadean type, and display an exarch condition not only in the 
leaf trace but throughout the petiolar system. 
The double trace is clearly present in the endarch series, and is 
also found in most of the exarch forms, including the higher Cycado- 
filices; Poroxylon, Cordaites, and their allies ; many modern coni- 
fers, and the living Cycadales. It seems clearly to have been a 
primitive character, and to have arisen from a constriction of the 
single diarch strand which was present in the ancestors of all seed 
plants. 
The endarch line of development apparently ended blindly 
and did not give rise to the typically endarch higher plants, which 
were developed along several distinct lines from a plexus of forms 
in the exarch group possessing a double leaf trace and seeds 
tesembling those of Medullosa. 
The Poroxyleae, Pityeae, and Cordaiteae came from this plexus 
at a very remote period. The tendency among this series has been 
toward the development of a parallel veined leaf, a clearly gymno- 
spermous type of reproduction, and an endarch condition of the 
central cylinder, together with a wide zone of secondary wood. 
The parallel venation of the leaf has doubtless been derived from 
a pinnate condition, with a two-bundled exarch rachis, a structure 
which has persisted only at the base of the trace. The cordaitean 
type of leaf seems to have existed in the Cretaceous, as is shown 
by the occurrence at that period of such forms as Niponophyllum. 
The genus V ilssonia, possessing a simple leaf with a distinct 
midrib but with clearly parallel lateral veins, shows an intermediate 
Condition between a pinnate cycadean leaf and a parallel veined 
condition, and gives a suggestion as to how the latter may have 
been produced. 
The cordaitean type of reproduction is unknown in the Cycado- 
filices, but Tri gonocar pon, the seed of Medullosa, approaches closely 
the seeds of the Cordaiteae. 
The endarch condition of the central cylinder has been caused 
by the disappearance of the centripetal primary wood of an exarch 
stem consequent upon the great increase in bulk of the centrifugal 
secondary wood, to which the protoxylem has finally attached itself. 
A second line from the double-bundled Medullosa-like plexus of 
