1901] SOUTHWESTERN CACTACEAE 41 
but with some anastomosing. The lateral branches, which 
apparently start for the tubercles, break up and form a network 
through the tissue. In this species the bundles appear to start 
ina definite circle in cross section, in contrast to the undulate 
line in which they occur in certain other forms. There seems to 
be no interfascicular cambium, 
the bundles in their exceedingly — pelea 1 ee sess 
slow growth apparently push- Re q ) 
ing their way outward into the 
“succulent tissue, compressing it 
about them, and taking up the 
room it formerly occupied, in- 
stead of the growth of the 
medullary tissue keeping pace 
with the increase of the bundles. 
The development of phloem is 
exceedingly slight, consisting 
wholly of sieve tubes and com- 
panion cells, with no trace of 
thick-walled elements. 
Of the xylem more needs 
to be said. A bundle is repre- 
sented in fig. 4. The elements 
which compose it are but two, ie fees a sects “dt sie 
¢ spiraled ducts and tracheids pundle in Cereus Fendleri. 
with a peculiar flattened spiral, 
the latter being by far the more abundant. These alternate in 
time of development, showing what appear to be periodic changes. 
The very earliest portions of the xylem are composed of the 
ordinary spiraled ducts, irregular in size and position, and there- 
after a small layer of these cells appears at what is probably the 
beginning of each season’s growth. The flat spiraled tracheids, 
an element already mentioned by Ganong and Schumann, form in 
definite rows, each apparently the product of a single cambium 
cell, which, after cutting off the few irregular segments to form 
the spiraled ducts, apparently undergoes no more division by radial 
re 
5 
OI 
= 
iG 
Py 
LY 
Ss 
LOTS] 
OO)p 
(% 
SY 
Serie 
SAIN 6 
COOLERS TOKO 
a oy) 
"> Ca) le Oe 
foloIG ee ASO 
