Tissues in Certain Monocotyledons. 57 
bundles of the outer zone, it is rare for the whole bundle to 
be formed from a single radial series. Usually two such 
series take part in its formation, sometimes one row forms 
the median part of the bundle, while the two adjacent rows 
cut off cells which contribute to form its flanks; sometimes 
all three rows contribute equally, but this perhaps only 
happens when an anastomosis is to be formed. The phloem 
of each bundle is formed very early. A cell cut off on the 
inner side of the cambium divides up by two or three inclined 
longitudinal walls. Thus a little group of small cells is formed, 
which represents the future phloem. Almost simultaneously 
an inner cell of the same radial row divides up to form the 
proximal part of the xylem. Next, cells situated at the sides, 
and either derived from the same radial series or from adjacent 
ones, take up the divisions, and form the lateral portions of 
the future ring of xylem. Lastly, fresh cells are added by 
the cambium to the distal side of the strand, and these 
ultimately complete the xylem towards the exterior. Of 
course all the stages run into one another, but it appears to 
be the rule for the phloem to take the lead. 
We had the same question to face here as in Dracaena and 
Yucca ; how do the tracheides of the secondary bundles de- 
velope ? The answer is here also a perfectly definite one ; 
they develope by sliding-growth alone. The comparison of 
transverse sections is never conclusive by itself, but it affords 
valuable indications. We can say at any rate this much ; not 
more than about twenty elements, as seen in transverse section, 
are, on the average, formed in the desmogen-strand by divi- 
sion. We know that the average number of elements in the 
transverse sections of a mature bundle is fifty-seven, of which 
forty are tracheides. The extraordinarily tortuous course of 
the tracheides, which form a sort of twisted skein in which the 
isolated cells of the xylem-parenchyma are entangled, and by 
which the phloem is enveloped, also strongly suggests an 
origin by longitudinal growth in a confined space (see Fig. 15). 
But direct evidence is not wanting. In longitudinal sections 
isolated desmogen-cells are sometimes found which have grown 
