LEAF-ARCHITECTURE AS ILLUMINATED BY A STUDY OF PTERIDOPHYTA. 687 



bud. The latter is the more usual process. ... It seems probable that the 

 appearance of pinnation in the frequent pairing of the pinnse at the base of the 

 rachis-segment is a secondary modification." This analysis was based upon pre- 

 served material, and supported by anatomical observations. It is probably correct. 

 Moreover, it receives support from the facts now demonstrated for the juvenile leaves 

 of M. pectinata. It is in fact just the sort of instability of the sympodial development 

 seen in them which, if continued in the larger adult leaf, would lead to something 

 not unlike the leaf-construction of M. sarmentosa. 



In neither species of Matonia does the leaf-architecture coincide with that of 

 Gleichenia, but both originate from a like sympodial source. But the modi- 

 fications are different in them all. M. sarmentosa appears to be more like 

 Gleichenia than M. pectinata. For in both of them arrest of branches is a feature. 

 In M. sarmentosa short shanks of dichotomy bearing usually two limited pinnae 

 and an aborted bud are produced, though not with regularity, on an approximately 

 equal dichotomising system. In Gleichenia similar shanks, which may or may not 

 lose their apical growth, and a bud, are borne on a sympodially developed scorpioid 

 system. The juvenile leaves of Gleichenia provide the simplest type of this leaf- 

 construction. It is thus possible to interpret all of them as variants of that 

 sympodial dichotomy to which the juvenile leaves of the primitive Filicales so 

 clearly point. 



Dipteridinse. 



The adult leaves of most species of Dipteris are notable for their regular 



dichotomy, of which the shanks remain separate laterally in D. Lobbiana and 



quinquefurcata, but are webbed into two equal coherent lobes in D. conjugata 



(Plate-fig. B). Notwithstanding this simplicity of branching, there is an advanced 



reticulation of the veins of the type " venatio anaxeti." But no one can doubt that 



the venation of the adult leaf is based on primitive dichotomy, notwithstanding the 



small-meshed reticulation characteristic of relatively recent forms. In Matonia the 



ultimate venation is forked, and usually free, with few and irregular fusions, and it 



thus stands on a more primitive footing than Dipteris. The reticulation appears 



even in the juvenile leaves of Dipteris, but it does not effectively obscure the 



dichotomy. An early leaf, perhaps the cotyledon, of D. Lobbiana is shown in 



fig. 20, a, which is simple in form and shows no completed reticulations. The single 



median vein and four lateral branches which it shows would be readily referable to 



a scorpioid sympodium. But there is no evidence, beyond comparison with Matonia 



and Gleichenia, that this is actually its character. Fig. 20, b, c, shows juvenile leaves 



which are bifurcate both in form and venation. Accessory veins already forecast the 



mature reticulation. Later leaves fork again (fig. 20, d), and so the plant advances 



towards the mature state.* Similarly with D. conjugata, notwithstanding the 



extensive webbing and profuse reticulation of the adult leaf, the juvenile leaf shows 



*■ Compare Sewakd, Phil. Trans., B, vol. 194, pi. xlviii. 



