BOTRYOPTERIDEAE 345 



represented the stock from which some at least of the 

 existing families of Ferns were subsequently derived. 

 Mr. Kidston has recently made a similar suggestion, 

 regarding the descent of Leptosporangiate Ferns from 

 the Botryopterideae as much more probable than from 

 any other Palaeozoic group so far as known to us at 

 present. 1 Mr. Arber, 2 taking a somewhat broader view, 

 derives the Leptosporangiatae from an ancient Palaeozoic 

 race, which he names the Primofilices, of which the 

 Botryopterideae were but one important family. He 

 further suggests (loc. cit. p. 227) that possibly the origin 

 of the Eusporangiatae is also to be sought for in the group 

 termed the Primofilices. I regard Mr. Arber's views as 

 an advance on those previously held. The Botryo- 

 pterideae, or at least their better -known members, 

 appear too specialised to have themselves been the 

 ancestors of the modern Ferns, but the analogies which 

 they present with so many families of the latter indicate 

 that they are probably an offshoot from the same main 

 line of descent. Kidston and Gwynne-Vaughan, in 

 discussing the comparative anatomy of Osmundaceae 

 and Botryopterideae, state that they " regard the 

 Osmundaceae as directly descended from an ancestral 

 stock from which at least two other types of structure 

 arose — that of Botryopteris and that of Zygopteris" s 

 a conclusion entirely in harmony with the general view 

 indicated above. 



1 " On the Microsporangia of the Pteridospermeae, " Phil. Trans. Roy. 

 Soc. B, vol. cxcviii. 1906, p. 442. 



2 "On the Past History of the Ferns,'' Ann. of Bot. vol. xx. 1906, 

 p. 221. 



3 R. Kidston and D. T. Gwynne-Vaughan, " On the Fossil Osmun- 

 daceae," Part i. Trans. Roy. Soc. Edinburgh, vol. xlv. Partiii. 1907. 



