Rendle. — On the Vesicular 
1 70 
that a similar development is common to all the genera where 
‘ vesicular-vessels ’ are found ; i. e. that they are formed from 
rows of cells by absorption of the transverse walls. Investi- 
gation shows, however, that in the onion each segment origin- 
ates from a single cell which is at first not very much longer 
than the cells of the neighbouring parenchyma, but dis- 
tinguished from these by its very large oval nucleus, and by 
its contents staining more deeply with iodine and the aniline 
dyes, especially Hofmann’s blue. This is well seen in tan- 
gential and radial sections of the base of an internal succulent 
leaf which has not yet begun to grow out (Fig. 1). 
Rows of somewhat elongated cells run longitudinally, 
parallel to the epidermis and separated from it by usually 
two layers of parenchyma. The transverse septa are not yet 
pitted, and sections treated (after Gardiner) for several seconds 
with strong sulphuric acid containing methyl-violet show no 
signs of communication between the members of the series, 
through the swollen cell-walls. With the outgrowth of the 
leaf from the bud the cells become proportionately elongated, 
and the transverse walls soon assume the usual pitted ap- 
pearance figured by Hanstein. In the succulent base of an 
elongated leaf the cells remain comparatively short, and at 
the extreme base are very short, and here the longitudinal 
series are irregularly connected, as Hanstein states, by cross 
unions (Fig. 6). In this succulent leaf-base they are almost 
invariably separated from the epidermis by two layers of 
parenchyma, rarely by one only ; in the green tubular part 
of the leaf, however, they often lie a little deeper, being just 
below the chlorophyll-containing palisade tissue, i. e. between 
the more special assimilating tissue and the elongated cells 
containing comparatively few chlorophyll-corpuscles between 
wdiich the vascular bundles run (Fig. 4). Here too the latici- 
ferous cells are much elongated. 
In the bulb-scale, as Hanstein states, two series are 
frequently found running alongside each other, and the wall 
separating them longitudinally is then also pitted ; when 
parenchymatous cells abut on them the dividing wall is simple. 
