DIV. i MORPHOLOGY 105 



the middle of the bundle. If the meristem is completely used up in 

 this process a closed collateral bundle results ; if some remains between 

 the xylem and phloem the bundle is an open one. In concentric 

 bundles the development does not follow a single type, and in 

 accordance with this the position of the protoxylem and protophloem 

 is various. 



Bundles in which the protoxylem is situated at the inner margin of the xylem 

 (in collateral bundles) or in the centre, as is often the case in concentric bundles, 

 are termed endarch. When the protoxylem elements are at the outer margin of 

 the xylem. as in radial bundles, it is spoken of as exarch. When the protoxylem is 

 in one or more groups removed both from the inner and outer margin of the xylem 

 it is mesarch, e.g. in the petiolar bundles of the Cycadeae or in concentric bundles ; 

 the protoxylem in this*case is embedded among the wider vessels. 



It is not at present known what relation holds between the arrangement of 

 xylem and phloem and the requirements of conduction in the plant, and whether 

 any one of the three types of bundle, e.g. the collateral, is superior in this 

 respect ( ). 



The phylogeny of the types of bundle is also not clear. All the evidence points 

 to the assumption that a stem with a single central vascular bundle is relatively 

 primitive. Such a bundle is found in the stems of a number of living and extinct 

 Pteridophyta and in all roots. The simplest and phylogenetically oldest type of 

 vascular bundle appears to be the concentric bundle with a solid central strand of 

 xylem ; at least this appears to be present in the young plants of nearly all existing 

 Ferns. The radial bundle also may be a very ancient type, as is suggested by its 

 constancy in the roots of all living and extinct cormophytes so far as our knowledge 

 extends and in the stems of some cormophytes. No other type of bundle is found 

 in both stems and roots. The variety as regards the construction and arrangement 

 of the bundles, which is met with in the shoots of Pteridophyta as contrasted with 

 the Spermatophyta, leads to speculations upon the mode of origin of these various 

 types of construction from stems with a single concentric bundle. There are stems 

 in which the vascular tissue of the single central bundle has the form of a hollow 

 cylinder enclosing a central strand of parenchyma or pith (Gleicheniaceae, 

 Schizaeaceae). In others the hollow cylinder of xylem is lined with an internal 

 zone of phloem (e.g. Marsilia). Lastly, there are cases in which the hollow 

 vascular cylinder is perforated by rhombic leaf-gaps at the departure of the leaf- 

 trace bundles (e.g. Aspidium filix 'mas}. In this last case a cross-section of the 

 stem shows a number of typically constructed concentric bundles, with solid central 

 strands of xylem, arranged in a circle. There are also forms in which a cylinder 

 of xylem immediately surrounding the pith is divided by radial plates of parenchyma 

 into a number of longitudinally-running strands of xylem placed side by side, the 

 whole being surrounded by a continuous zone of phloem (e.g. Osmunda). Lastly, 

 there are cases in which the phloem is correspondingly divided so that the radial 

 plates of parenchyma separate, as medullary rays, the collateral strands composed 

 of xylem and phloem (e.g. rhizome of OpMoglossum}. These examples show how 

 either a reticulate tube of concentric bundles or a hollow tube composed of 

 collateral bundles can be derived from a centrally-placed concentric bundle. If we 

 assume that the phylogenetic development has proceeded on these lines, it is 

 clear that neither one collateral bundle of the Spermatophyta nor one of the 

 circle of concentric bundles found in many Ferns is homologous with the central 

 bundle of " primitively constructed " Pteridophyta. The totality of collateral or 



