IN CEYLON. 
65 
three portions, together with a prolongation of epicotyledo- 
nary strands to different distances in the primary root. 
In seedlings where the epicotyledonary traces are not 
continued into the primary root a similar splitting of the 
xylem of each cotyledon trace may occur. 
The presence of twelve to twenty-two xylem strands in a 
comparatively narrow cylinder, where each strand is relatively 
large, renders the actual behaviour of each strand difficult 
to follow. 
The reason for such profuse scattering of the proto-xylem 
system is perhaps to be found in the bulky nature of the 
parenchymatous system. In all cases the scattered condition 
attains its maximum in the collet area where the parenchy¬ 
matous system is at the maximum, and it would appear 
probable that the mechanical support and supply of nutrition 
were the objects aimed at. The splitting is always more 
complicated in a wide parenchymatous seedling, such as D. 
Embryopteris, than in a narrow one, such as D. Ebenum, and 
the fact that the strands are thus distributed prior to the 
appearance of a vascular cambium is sufficient proof that the 
diffuse condition of the proto-xylem is not due to the pressure 
of new cambial products. 
Direction of Lignification and Differentiation of 
the Vascular Elements. 
If the embryo or young seedling be examined the differen¬ 
tiation of lignified elements can be determined. We may 
assume that lignification commences in the oldest element 
first and gradually appears in successively younger elements. 
that the direction of differentiation of the elements is the 
same as that of lignification. In the embryo the beginning 
of lignification is seen in the xylem of the cotyledonary traces, 
and commencing at the cotyledonary node the lignified 
elements can be traced downwards into the primary root and 
upwards into the petiole of the cotyledons. The lignification 
at this stage is somewhat irregular, and appears to be a little 
8(1)4 (9) 
