XXIII] COMPARISON 187 



use here should lead to some definite conclusion whether or not the suggested 

 relationship is a real one. 



(i) The rhizomatous shoot of the Marsileaceae, with alternate leaves 

 and lateral branchings which suggest a reference to a dichopodial branching, 

 find their correlative in Lygodiicni and in the creeping species of Anemia 

 (^Anemiorrhizd). But it would also compare with rhizomes of many other 

 creeping Ferns. 



(2) The apical meristems conform to what is usual in Leptosporangiate 

 Ferns: even the segmentation of the young leaf is the same with its marginal 

 series of cells. 



(3) A comparison of the leaves within the Family shows successive steps 

 from a simple unbranched subulate leaf to those with two, four, or even six 

 pinnae. The venation of these pinnae is clearly dichotomous, and the equal 

 forking is obvious in the last branchings. But all of them are found on 

 exact analysis to be referable to dichotomy developed sympodially (com- 

 pare Bower (471), Figs. 8, 9). This conclusion coincides with the result 

 of analysis of the primordial leaves of Osnmnda, Botrychium, Anemia and 

 Lygodium, and may be regarded as a general feature in relatively primitive 

 Ferns. Schizaea itself illustrates the more primitive equal dichotomy in its 

 sterile region: but in the fertile region of its leaf the branching is dichopodial 

 {I.e. Fig. 6 bis). A feature of advance in the leaves of Marsilea and Regncl- 

 lidium is the marginal anastomosis of the veins: but in advanced species 

 of Lygodium and of Anemia vein-fusions are present. The leaves of the 

 Marsileaceae may then be accepted as being originally constructed on a 

 plan not unusual in relatively early Ferns : that plan is seen most obviously 

 in the adult leaves of Marsilea, with four or sometimes six pinnae : but it is 

 simplified in Regnellidium to only two pinnae, while in Pilulai-ia the leaf 

 is unbranched. This state is seen in the juvenile leaves of Marsilea itself, 

 and is exemplified also in the sterile blades of many species of Schizaea, 

 and particularly in .S. nipestris. 



(4) The solenostele of the rhizome,and the undivided leaf-trace of the Mar- 

 sileaceae find their correlatives in various primitive Ferns, in particular the 

 details show some near similarity in the creeping species of Anemia included 

 in \Anemiorrhiza. In Pilnlaria the inner endodermis ma}' be absent: and 

 thus its stele approaches that simple state seen in some of the smaller 

 species of Schizaea (Chapter XXII, Fig. 443). Lastly, the marginal origin of 

 the pinna-trace, which may itself be held as indicating the pinna-nature of 

 the sporocarp, corresponds to its marginal origin in the Schizaeaceae: it is 

 found to apply in M. polycarpa for each of the numerous sporocarps (Fig. 

 462). 



(5) The dermal appendages of the Marsileaceae are simple hairs, as is 

 commonly the case in primitive Ferns, and in particular in the Schizaeaceae, 



