444 MR. F. O, BOWER ON THE STRUCTURE OF 
the bundles of the leaf-trace, are found only in comparatively few 
cases; and in many of those cases such cortical bundles may be 
regarded merely as branches or continuations of the common 
bundles of the leaf-trace, which ultimately enter the normal ring. 
This is the case in species of Lathyrus, in Casuarina, and many 
Begonias. In other instances the cortical bundles have a similar 
origin to those above cited ; they, however, do not enter the ring, 
but form a cortical system connected with the ring by anasto- 
moses at the nodes. This arrangement is found in the Caly- 
canthee and many Melastomacee. In a third series of succulent 
plants, including especially forms with reduced leaves, bundles 
are found ramifying in the cortex as in the lamina of the leaf; 
and, finally, in the winged Rhipsalidacez the common bundles of 
the leaf-trace are chiefly cortical and surround a central ring, 
which is for the most part cauline. 
It is clear that the case of Rhynchopetalum does not coincide 
exactly with any of these, since, as above stated, the cortical 
system does not consist of branches of bundles of the leaf-trace, 
but are cauline bundles. A closer comparison may, however, be 
drawn between the stem of Rhynchopetalum and that of Cycas 
as described by Mettenius *. There the girdle-like bundles of 
the leaf-trace are connected with one another by bundles, which 
pursue a nearly vertical course and together form an *' accessory 
cortical system." They originate from a secondary meristematic 
activity of longitudinal rows of cells of the cortex, strands of 
small cells being thus formed, which develop into small vascular 
bundles with a radial arrangement of their elements round 4 
centre. This is fundamentally the same process as has been 
above described for the cortical bundles of Rhynchopetalum. 
The chief difference between the two cases lies in this : that in 
Rhynchopetalum the cortical bundles run obliquely, and together 
form an approximately regular network with four-sided meshes, 
which bear a definite relation to the bases of the leaves, and 
therefore also to the bundles of the leaf-trace ; in Cycas, how- 
ever, the bundles of the accessory cortical system are not thus 
regularly arranged, and pursue an almost vertical course. In 
both cases the mode of origin is the same ; and in both cases the 
* On Cycas circinalis, cf. description by Miquel, ‘ Ueber d. Bau’ &c.; 0n 
Cycas circinalis, Linnea, Bd. xviii. p. 125. On Cycas revoluta, cf. Mettenius, 
“ Beitr. z. Anat. d. Cycadeen,” Abbandl. d. k. Sächs. Ges. d. Wissensch. vit. 
p. 567. 
