CEYLON. 
In some specimens of D. raontana from Trincomalee I 
observed that some oblique bands of wood parenchyma were 
connected at the outer end to the median cells of a tangential 
band of parenchyma, and within, to a medullary ray at a 
point opposite to another tangential line of parenchyma. 
By this means tangential bands of one zone can be con¬ 
nected, almost directly, with a band in an older zone, along 
a different radius. Such an arrangement must greatly 
facilitate the transference of food materials through the 
parenchymatous system. 
The tangential and oblique bands of wood parenchyma 
therefore effect an excellent communication with all 
elements, excepting perhaps the vessels and tracheids. The 
remaining part of the wood parenchyma is, however, specially 
associated with these elements. In most of our species the 
large vessels are surrounded or flanked on their tangential 
surfaces with relatively wide lumined but short wood- 
parenchyma cells. These are in communication with the 
other cells of wood parenchyma or medullary rays, and 
therefore a complete communication is effected with every 
element of the secondary xylem. 
The vertical distribution of these elements is easy to 
follow since they are arranged one above another through 
a considerable vertical distance. They form very long 
columns of cells, often as many as twelve cells in the same 
row. The majority form single straight lines with the 
components exactly above one another; frequently one cell 
abuts on to another, not end to end, but 1>y the radial 
walls, thus giving the terraced system in the vertical 
direction. 
There is nothing unusual in the form of these elements. 
In transverse outline they may be any shape from circular 
to crudely four-sided, but only rarely do they exhibit sharp 
corners. They are elongated considerably in the vertical 
direction, the vertical walls being straight and the end walls 
flat or occasionally tapering at one end when in contact 
