242 BOTANICAL GAZETTE [MARCH 
other herbaceous plants studied. In Cucurbita maxima, for 
example, as many as 9g lateral callus pads have been seen in a 
continuous row in a single sieve-tube. 
__ The evolution of the sieve-tube parallels that of the tracheid or 
vessel. For example, in Ephedra the vessels have oblique end walls 
with several round or oval pits. This is supposed to be the primi- 
tive type of vessel. In Liriodendron the end wall of the vessel is 
still rather oblique, but the pits are of the scalariform type and 
close together. In the highest type, the pitted vessel, the little 
margin that is left of the end wall is usually at right angles to the 
side walls, as in Fraxinus. 
Conclusion 
In studying the phloem of the dicotyledons, it has been found 
that there is a gradual transition from the gymnospermous type 
of sieve-tube to the so-called dicotyledonous type as seen in the 
Compositae. At the first stage in advance the lateral plates are 
smaller and have smaller meshes than the terminal plates; then 
as the terminal wall becomes more and more at right angles to the 
lateral walls, the number of terminal plates decreases until there 
is only one terminal piste with relatively large meshes, and the 
lateral plates become “‘sieve-fields” or lattices. 
Paleobotany, ontogeny, and studies of xylem have induced 
many botanists to believe that herbaceous plants are more advanced 
in their evolutionary development than woody plants. This 
study of the sieve-tube adds another argument in favor of this 
view. 
The first two years of this work were done in the Phanerogamic 
Laboratory of Harvard University under the direction of Professor 
E. C. Jerrrey, and the last year of work has been under the direc- 
tion of Professors J. M. Courter and W. J. G. Lanp at the 
University of Chicago. The writer wishes to express his thanks 
to these instructors for their helpful advice. 
TRANSYLVANIA UNIVERSITY 
LExIncTon, Ky. 
