LEAF- ARCHITECTURE AS ILLUMINATED BY A STUDY OF PTERIDOPHYTA. 703 



(in) where the development of the leaf is strong there may be transition to the 

 Monopodial Branching. 



VIII. But in the individual Fern-leaf the order of events is developmentally the 

 reverse of this. The first-formed branches (pinnae) are monopodial, the later of 

 sympodial origin, and the final branching dichotomous. Thus the distal end of the 

 Fern-leaf may be held to retain its primitive characters. Anatomically it is found 

 that ancestral features are also located in the leaf-trace at the base. It is thus 

 possible to regard the middle region of the leaf as an evolutionary innovation, a 

 view which accords equally with anatomical fact and with the comparative study 

 of leaf-development. It may then be stated as a working hypothesis that the most 

 archaic regions of the leaf (especially in Ferns) are the apex and base, and the 

 middle region is derivative. 



IX. In some cases of highly organised leaves, especially in Ferns, the branching 

 of the adult leaf may be throughout that of equal dichotomy. This is usually 

 accompanied by a long petiole, so that the orbicular or laciniate lamina is expanded 

 far from the leaf-base. Historically this may be either a retention of the primitive 

 dichotomy or a reversion to it : an opinion can only be formed on the facts for each 

 individual case (figs. 27, 28). Thus, in some highly organised leaves the sympodial 

 or monopodial states may not be represented at all. 



X. In other cases the converse may happen, as in the adult leaves of Angiopteris* 

 where, though in the young leaf the branching is a scorpioid sympodium, the 

 relatively few pinnae of the adult arise monopodially, and the phyllopodmm is then 

 arrested in its growth. It does not progress to the primitive dichotomous stage, 

 which is thus omitted from the adult leaf. A similar state is seen in the Cycadaceae.t 

 In them even the acropetal succession is not strictly maintained. This probably 

 gives a correct clue to the architecture of the leaves in most Angiosperms, in which 

 arrest of apical growth at an early stage is a marked feature of the leaf-development. 

 Certain extreme examples of such arrest of apical growth involve the atrophy of the 

 whole distal region. They are seen in simple form in the protective scales of the 

 Osmundacese, where the whole branch system of the lamina is arrested in its early 

 stages, while the sheathing base remains of normal size. The same is seen in the 

 Cycadacese, In Angiosperms this is the case with bud-scales, in which in various 

 degree the "lamina" is represented in atrophied state, while the intercalary petiole 

 is not initiated.! Such cases may be held as further advances in that arrest which 

 is already seen in Angiopteris. Not only is the phyletically earlier, dichotomous 

 region of the distal branching omitted, bjut the whole of it ; and the effective part of 

 the scale originates from the " hypopodium" (Blattgrund of Eichler). Thus it may 

 be stated that in some Ferns, in Cycads, and in almost all Angiosperms the apical 

 growth of the leaf-primordium is arrested early, and the distal primitive stage of 



* Phil. Trans., pt. ii, 1884, rigs. 15-17. t Ibid., pt. ii, 1884, figs. 24-44. 



I Goebel, Organography, Engl, ed., ii, p. 386. 



