444 
Hill and de Fraine . — On the 
wood has been becoming gradually less, especially the metaxylem, so that, 
in transverse section, it is impossible to distinguish with certainty the proto- 
xylem from the metaxylem elements. 
That part of the cylinder marked p . $ now becomes detached and joins 
on to the opposite horn ; in the gap thus produced a group of xylem 
elements is left in situ , and forms one pole of the root-structure (Diagram 3, 
Fig. 6). At a slightly lower level the phloem in the region between c and d 
divides, exposing another small mass of wood which forms another pole of the 
