THE VASCULAR SKELETON 689 



within the cortex (Cheirostrobus), giving sometimes a median bundle 

 (Ophioglossuni), sometimes a paired trace (Bofrychii/tn). All the more 

 primitive types of Ferns, including the fossil Psaronius, have a single 

 more or less horseshoe-shaped trace ; but the modern Marattiaceae and 

 the bulk of the Polypodiaceous Ferns have a trace composed of many 

 strands : these are, however, arranged in series corresponding to the 

 horseshoe outline of the undivided trace. The facts indicate with no 

 possible uncertainty that there has been a disintegration of the leaf-trace 

 by fission : it finds its origin in branching of the strands in an enlarged 

 upper region of the leaf, and has been phyletically progressive from a region 

 lying above towards the base. A comparison of Fig. 97 will make this clear : 

 leaf-traces are there shown each of which consists at the base of a broad 

 strap-shaped strand : this breaks up distally into numerous strands. But in 

 Cyathea, which is structurally a more advanced type, the breaking up has 

 been continued down to the base, and the leaf-trace comes off initially as 

 numerous separate strands (Fig. 337). The same has probably happened 

 in the modern Marattiaceae as compared with Psaronius ; in most Mixtae 

 as compared with the Gradatae (p. 648), and in the section Ophioderma as 

 compared with Euophioglossum (p. 462). Thus in several distinct phyla 

 it is shown that a progressive disintegration of the leaf-trace has been 

 effective ; and goes always with leaf-enlargement just as disintegration of 

 the axial stele has followed on expansion of. the axis. But in both cases 

 the enlargement has phyletically preceded the consequent disintegration. 1 



The present interest in these complex structures in axis and leaf-stalk 

 does not lie in their detailed study, so much as in the fact that in all 

 cases they appear only in the plant when advanced towards full develop- 

 ment In the young seedling a stelar structure, little removed from or, in 

 most cases, actually showing a protostelic state, is constantly found ; and 

 from it the various steps may be traced to the more complex condition. 

 So far as the development of the individual can be held to reflect the 



1 In certain Pteridosperms and Gymnosperms a double lenf-trace has been found lo be 

 prevalent, and it has been suggested that it finds its origin in the bifurcation ol tin- leaf. 

 Arguments based on the existence of a double leaf-trace have recently been extended to 

 Flowering Plants (Miss Thomas, New I'hytologist, 1907, p. 77). It is nut proposed here to 

 criticise those arguments, but merely to point out from the .side of the Pteridophyta that 

 there is no constant relation between foliar dichotomy and a double leaf-trace. In Sigillaria, 

 Kidston (Proc. R.S., Edin., vol. xxvii., p. 203) has shown that the double leaf-trace, 

 already recognised by Renault, exists in a leaf of simple-form ; on the oilier hand, the 

 bifurcate sporophyll of Tmesiplcris has only a simple leaf-trace. In the Ophioglossaceae, 

 Eiiophioglossiiiii ami ll<-lmiiith,<.,l.i- ,'iys have a simple leaf-trace, which soon biauches, 

 Botrychinin has a double leaf-tree, Ophioderma a trace of several strand>, not arranged 

 in any binary scheme (Ann. of Bot., xix., I'l. xv., Figs. 6-29). Lastly, in many primitive 

 Ferns, where dichotomous and other branching of the leaf is prevalent, tl e is a 



single strand. Such facts suggest the propriety of extreme caution in applying urgum 

 based on the vascular structure at the base of the leaf. It would seem not improbable that 

 a double leaf-trace might appear in any broad flattened organ which is bilaterally sym- 

 metrical, whether branched or not. This may very well have been the case in Si & liana. 



2 X 



