dles." We have seen that a given strand or bundle may run for part of its course simply side by side with another and separate from it; at other parts of the course the bundles may be united with others. In the case of the oak it will be clearly borne in mind that the individual or separate bundles of the leaf-trace pass into the stem at the node of insertion of the given leaf, and then run down side by side at a practically constant distance from the surface of the epidermis on the one hand, and the longitudinal axis of the pith on the other. At different levels below, at or very near nodes, these bundles turn aside laterally—i.e., in the tangential plane, and hence, still keeping their mean distance from the epidermis and pith, join with others.
This being understood, it is also obvious that on the whole the collection of vascular bundles in a young branch form a nearly cylindrical trellis-work or mesh-work symmetrically disposed between the pith and the cortex, and that the latter (cortex and pith) are in connection through the meshes between the interpectinating and concomitant vascular bundles. These radial connections of the pith and cortex are the primary medullary rays.
It will now be clear why we observe on transverse sections of the young stem taken across an internode the arrangement shown in Fig. 9. The vascular bundles are grouped in a ring round the pith, separating it off from the cortex and its covering the epidermis, and with