304 
EDMUND W. SINNOTT 
conservatism. The structure of the angiospermous petiole, indeed, 
has been the subject of numerous and careful studies by different 
anatomists, and it has been found that in a few families and even in 
certain genera the petiolar structure is sufficiently peculiar and con- 
servative to be used as a diagnostic character for the group. In 
general, however, the number and arrangement of the petiolar bundles 
is too much dependent on the size and texture of the leaf to be of very 
great taxonomic importance; and even where these characters are 
reasonably constant, it is very hard to draw phylogenetic conclusions 
from them and to determine which particular type is the most primi- 
tive. The fact that the node of several of the lower orders, rather 
than the petiole, seems to be a notably conservative region suggested 
that in the higher seed plants, as well, it might be worthy of study. 
The present investigation has therefore had as its object an examina- 
tion of the nodal structures of the angiosperms with the hope of dis- 
covering simple and constant anatomical features which will be of 
value in determining broad lines of relationship. 
In the Lycopodiales, ancient and modern, the vascular supply for 
the leaf is at its base almost invariably a single strand. In the ferns, 
cycads, Cycadofilices, Cordaitales, Ginkgoales and conifers the foliar 
bundle, although often single, seems primitively to have been a double 
one. This double trace, especially in the ferns, has often been broken 
up into a wide arc of strands. In these lower vascular plants the 
foliar supply, whether single, double or multiple, causes typically but 
a single gap (if any) in the continuity of the vascular ring {fig. i). 
Thomson (7) has observed, however, that in very vigorous specimens 
of Agathis the two bundles of the leaf-trace are separated at their 
insertion by a tiny segment of the secondary wood and thus cause a 
double gap in the cylinder {fig. 2). Among the Gnetales the genus 
Ephedra, which most closely approaches the conifers in other respects, 
has like them a double leaf -trace, but the two portions as in Agathis 
are separate at their insertion {fig. 3). The node of Gnetum is much 
more complex for here there are from seven to eleven strands passing 
off into each leaf and every strand causes a gap of its own in the 
cylinder. 
Such a condition where there is an odd number of bundles, each 
departing from a distinct gap in the vascular ring, is typical of the 
Angiosperm node, and has apparently arisen as an adaptation to the 
increased transpiration current passing to the broad leaves. The 
