366 
L. H. MACDANIELS 
parenchyma cells proper. This gives rise to a vertical series of cells 
with the cell at each end of the series pointed, as was the cambium 
from which it was formed (figs, ii, 32). Parenchyma of the true 
cambiform type would be produced either by the failure of the paren- 
chyma mother cell to divide transversely or through its division by 
oblique cross walls so that resulting cells would be prosenchymatous in 
shape. Such division, however, was not observed, the cross walls 
being nearly transverse in all cases. The divided-cambiform type 
shows great variation in shape and size, from elongate, narrow cells 
with little radial depth, to nearly cubical cells approaching the con- 
ducting type. 
The pitting in the divided-cambiform parenchyma is in most cases 
very prominent on the radial and terminal walls. Frequently the 
pits are grouped so as to present a sieve-like appearance as shown in 
the parenchyma of Corntis paniculata (figs. 37, 38), Castanea dentata, 
and others. In some species the pitting is not so pronounced, but the 
entire walls of the cell are thin and probably allow rapid diffusion. 
These types of parenchyma, according to Haberlandt (4), serve in the 
conduction of carbohydrates and the more easily diffusible proteins, 
a function for which they are apparently well suited histologically. 
As before indicated, the present study does not confirm the phylo- 
genetic significance of sieve-tube type as maintained by Hemenway 
(6). In making such a statement, of course, it must be borne in mind 
that in evolutionary development morphological characters do not 
advance equally. A given group may thus have one character highly 
developed and yet possess others remaining at a low level, as, for 
example, the presence of the inferior ovary, a "high" character, in 
"low" families such as the Juglandaceae and the Hydrocharitaceae. 
It must be recognized, also, that the number of species in this research 
is perhaps too small, as compared with the vast number of woody 
plants in the world flora, to justify the making of any sweeping general- 
izations. It is particularly unfortunate, too, that there are not in- 
cluded representatives of the tropical floras where woody plants form 
so large a proportion of the whole. Yet, inasmuch as the material ex- 
amined was selected to represent, in so far as possible, a series of woody 
plants from the lowest to the highest forms in the Engler system, and 
since references in literature to the phloem of tropical woody plants 
show that the condition found there is not different from that in the 
