386 Darhishire. — Observations on Mamillaria elongata. 
nature is. It is important, however, to notice that the one branch from the 
main bundle divides, but that its branches in the tubercle and also in the 
cortex anastomose freely, and in the tubercle itself finally end in a large 
cup-like mass of big tracheids (PL XXV, Fig. 15). 
On leaving the main bundle the lateral branch leading to the tubercle- 
system consists of extremely minute spiral tracheids. They are 13 to 1 6 pc 
in diameter. In the protoxylem the spiral bands are about 2 ju, thick and 
as much as 8 fx apart. In the younger metaxylem the figures would be 
2-5 and 3-4^ respectively. I was unable to make out the length of the 
tracheids at this point in the bundles. But they are evidently fairly long, 
at any rate in proportion to their diameter. In this they differ very much 
from the spiral tracheids of the metaxylem of the main bundle. They are 
in fact more of the type of spiral vessel or tracheid met with in most 
protoxylems of the normal Angiosperm. Here and there we do, however, 
get one of the tracheids in the cortical bundles suddenly passing into 
a large spiral tracheid, measuring as much as 80 by 36 [a, which is of the 
typical cactaceous form. It will be found to be in contact with some two or 
three large parenchymatous cells of the cortex. 
The bundles branch fairly frequently and anastomose again freely. 
A certain number of branches pass towards the growing-point, which is 
situated in the axil of the tubercle, but a greater number of bundles bend 
out and grow towards the outer end of the latter. Of these a certain 
number pass towards the rows of palisade-cells of the wart, and here they 
end blindly (PI. XXV, Fig. 15). 
All the bundles are accompanied by bast, the position of which, how- 
ever, changes in such a way that finally it always lies outside the wood. 
The bundles passing to the tubercles therefore contain the following 
structures : bast, small and narrow tracheids with very close spirals, and 
larger tracheids with very loose spirals. This is what one might expect, 
but owing to the short distance between protoxylem and bast the contrast 
between the two forms of spiral tracheids is very marked (PI. XXV, Fig. 19). 
The bast may extend to as great a depth as the wood when entering the 
tubercle, but later on it becomes very much reduced. The very last 
endings of the central mass of vascular tissue just underneath the cushion 
of spines are quite free of bast, being surrounded by parenchymatous 
cells only. 
In a transverse section of a tubercle we would be able to notice, when 
cut half-way down, about 6-7 bundles which form an outer cortical ring 
(PI. XXV, Fig. 10). These bundles end just inside the palisade-tissue. 
At the first point the cortical bundle would generally be about 70-90 fx 
deep in a radial direction, and about half that in a tangential direction 
with regard to the periphery of the whole tubercle (PI. XXV, Fig. 7). 
In bulk the wood ( b and c) is slightly in excess of the bast, although the 
