LEAF-ARCHITECTURE AS ILLUMINATED BY A STUDY OF PTERIDOPHYTA. 701 



the closely related Osmunda regalis (fig. l), while Todea barbara (fig. 5) takes a 

 middle position. Other examples may be found among the species of Hymeno- 

 phyllum, of Botrychium, and of Dipteris. In fact, there is hardly any large genus 

 of Ferns in which the gradations of " webbing " of the branched leaf between the 

 laciniate and the webbed are not found. Sphenophyllum illustrates the same pheno- 

 menon ; for the archaic S. trichomatosum has deeply cut leaves, while C. cuneifolium 

 has its leaves entire, with a slightly wavy distal margin, and Trizygia has even the 

 margin entire, and the leaf expanded as a large assimilating organ. Such facts 

 suggest that a progressive " webbing " of leaves originally laciniate has taken place 

 within the family. The facts for the genus Dipteris * are hardly susceptible of 

 any other interpretation, and there is reason to believe that it occurred extensively 

 among early Vascular Plants. 



Already Potonie f has indicated that in Sphenophyllum the species from the 

 lower horizons are laciniate, while those of the more recent are entire, indicating a 

 progressive webbing. To this view for the Sphenophyllacese Dr Kidston assents. 

 But, however cogent the evidence of this may be for certain cases, there is no doubt 

 that in the course of descent the reverse has also happened. It is illustrated in the 

 submerged species of Ranunculus and Potamogeton. The general conclusion will 

 then be this : that it is probable that in leaves with branched venation a progressive 

 webhing has occurred in the course of evolution. But where there is an actual 

 state of laciniation seen in any given case, it may either be really the primitive 

 state preserved, or an adaptive reversion to it,\ 



V. The dichotomy in juvenile leaves, whether as seen in their laciniate outline 

 or only in their venation, may be equal or unequal. In the latter case the system is 

 commonly developed sympodially, and all possible gradations may be observed from 

 the one to the other. These are well illustrated in the juvenile leaves of Pteridium 

 (fig. 26) and of Osmunda (fig. l). It is important to note that the order of the 

 ontogenetic progression is normally from equal dichotomy to sympodial development. 

 Potonie also points to a similar progression in the juvenile leaves of Flowering 

 Plants, § in which the outline may be dichotomous, though not so in the adult leaves. 

 He quotes Kny and Winckler to the same effect. But to take the mere outline 

 into consideration without the venation may be misleading. Apparent dichotomy 

 may be merely a result of deep emargination, and there are indications in the 



* Land Flora, p. 618, etc. f Lehrbuch, p. 176. 



I It may be a question whether all cases of complex venation in any leaf-expansion represent the result of 

 webbing of lacinipe originally distinct. It is possible that a branching of veins may be initiated ale novo in a leaf- 

 expansion. We know that some such new developments have produced the reticulate state. It would therefore be 

 rash to deny that something similar may in some cases account for the origin of additional veins in an entire blade. 

 We may conclude that in primitive types with open venation which are "webbed" the venation may be a near 

 index of a primitive laciniation. But in derivative types, and especially where the venation is reticulate, as ia 

 Angiosperms, this may have so obliterated the original scheme of construction that what is actually seen can no 

 longer be tiusted as indicating it further than in quite general features. 



§ Her. d. d. Bot. Ges., xiii, p. 245. 



