218 The Botanical Gazette. {May, 
of it. In what does ‘‘vitality” of this sort consist? Is it a 
more hardy composition of the protoplasm of that plant, or is 
it in the fact that nearly all cells are alive, i. e., non-skeletal, 
— result of the abundant water and food supply in every 
cell? 
The bundle systems are built upon a very uniform system; in 
the several genera examined they form modifications of that 
which occurs in Opuntia, which in turn is easily derivable 
from that of Peireskia. This should be examined for the remain- 
ing genera. The bundle systems seem to follow closely external 
form and morphological changes, a fact which comes out with 
particular beauty in Mamillaria, where the grooved and un- 
grooved forms have bundle systems answering exactly to the 
morphological difference between them. Further study of this 
relation between internal anatomy and external changes is 
needed before generalization is possible, but the point is as a 
principle one of great interest and importance. A careful study 
of the sieve elements is needed. Many thick forms of Cereus, 
Melocactus and Mamillaria show a pith-system of interlacing 
bundles, the development of which is still unknown. In 
Anhalonium, and confined to it, is found a remarkable bundle- 
cylinder system. It resembles somewhat, though in more 
extreme degree, that figured for the separate cylinders of 
Bauhinia.?° In cross-section the bundles radiate fan-like in 
clusters, and then curve about plume-like, so that they often 
come to face inward towards the center of the stem. Its 
development is unknown. The entire minute anatomy of 
The constancy of the behavior of the axillary vegetative 
points in the same ‘‘genus,” and the differences in its behavior 
in different “‘genera” have suggested that it may be used to 
indicate true relationships, and I have given in Flora (I. c-) @ 
scheme showing such relationship. This point needs much 
more full investigation. For systematic purposes, this family 
needs above all others exact study in the field, for variation 15 
rife in it. The affinities of the family as a whole are not 
*® See Engler & Prantl, Pflanzenfamilien. Leguminose, p. 81. 
