LEAF- ARCHITECTURE AS ILLUMINATED BY A STUDY OF PTERIDOPHYTA. (597 



situated on pairs of the parallel veins, as typical of Scolopeiidrium. The interest 

 of this species is first in comparison with S. vulgare ; the blade may here be held 

 to be a shortened and condensed form of that there seen. But the structure in 

 S. vulgare has recently been elucidated by comparison of Blechnurn punctulatum* 

 This very variable leaf has been shown to be a derivative from the typical Blechnurn 

 sporophyll ; the latter is a probable derivative from a Cyatheoid source, such as 

 Matteuccia, and ultimately with probability from a Gleicheniaceous affinity. But 

 Matteuccia has an ordinary pinnate leaf. Unless, then, all the comparisons given in 

 detail in my memoir on Blechnurn are at fault, the conclusion will be that the 

 reniform blade of S. Delavayi is the result of a condensation of the leaf, with 



Fig. 30. — Basal part of the lamina of an adult leaf of Scolopcndrium Delavayi, Fr. ( x 4.) 



simplifications of its outline, starting from a pinnate — that is, a scorpioid — type. This 

 case is important in its bearing upon "webbing," and indicates that here an entire 

 reniform leaf is probably derivative from a more elaborate and laciniate type of leaf. 

 Another peculiarity which is striking, but by no means restricted to any given 

 systematic group, is the unilateral branching of pinnae, by helicoid development, as 

 in Matonia and others. It is seen in Pteris semipinnata in a very prominent form 

 (Plate-fig. E). While Pteris quadriaurita and others show an equal lobation on 

 either side of the pinna, or pinnule, in P. semipimiata the branching appears on the 

 katadromic side only (fig. 31). This may be interpreted as the unequal development 

 of successive dichotomies, so that the anadromic shank is constantly the stronger. 

 A similar condition has been described for Stenochlsena by Christ.! On his plate iii 

 he represents it for a species from Borneo, and on plate iv for one from " Deutsch- 

 Neu-Guinea." Thus it is not a local peculiarity. This anadromic helicoid branching 

 is also represented in the fossil Odontopteris minor, Brongn., as shown by Zeiller 



* Ann. of Bot., vol. xxviii, pp. 405-411. f Verhandl. d. Schw. Naturf. Ges., 89, p. 178. 



TRANS. ROY. SOG. EDIN., VOL. LI, PART III (NO. 21). 102 



