178 D. H. Scott. 



does not include the Botryopterideae, whicli. as he himself recognizes, 

 are certainly represented in Lower Carboniferous rocks. This family, 

 in fact, has come to occupy an important position, for in the present 

 state of our knowledge it represents the best-attested group of 

 Palaeozoic Ferns, as will be shown below. Almost all the well-known 

 and striking genera of Fern-like fronds have now come under suspicion, 

 and cannot be accepted as alîording in themselves any evidence for 

 the existence of Ferns, as distinguished from Fern-like Spermophyta. 

 The presence of seeds has been actually demonstrated in members 

 of the genera Sphcnopteris, Neuropteris. Aneimites and Pecopteris; the 

 evidence is almost equally convincing in the case of Alethopteris-, in 

 numerous other genera, such as Eremopteris, Odontopteris, Linopteris 

 and LoncJiopteris, all the indications are in favour of seeds having 

 been borne, though at present there may not be much beyond associ- 

 ation to guide us. It is highly probable that some of these form- 

 genera are purely artificial associations, which include Ferns as well 

 as Fern-like Seed-plants; in Pecopteris especially, while one species, 

 P. PhicJieneti. undoubtedly bore seeds, as Grand'Eury has shown, 

 many others show the well-known fructifications commonly regarded 

 as those of Marattiaceous Ferns. Even in the latter cases, however, 

 the question is not free from difficulty, for recent work has proved 

 that some of the supposed Marattiaceous fructification were in reality 

 the microsporangia of Pteridosperms (Kidston, 1905). 



At the present moment it has thus become extremely difficult to 

 know what plants should be included under Palaeozoic Ferns. The 

 course I propose to take here is first to discuss the family Botryop- 

 terideae, which on present evidence appear to have been represen- 

 tatives of a comparatively simple type of Filicineae, of great antiquity. 

 Whether we call them "true Ferns" or not is immaterial; they 

 certainly do not belong to any of the existing families, though they 

 show relations to them in various directions. Next, a few outlying 

 forms will be referred to, which we may group, if we like, side by 

 side with the Botryopterideae, under the wider designation Primo- 

 filices, recently proposed by Mr. Arber.^) Finalh^, the evidence for 

 the existence of Marattiaceae in Palaeozoic times will be considered. 



Botryopterideae. 



The family was first founded by Renault, to whom our knowledge 

 of the group is primarily due. In the type genus, Botnjopteris, at 

 least two British species are now included, in addition to B. forensis 



^) E. A. Newell Arber, The Past History of the Ferus. Ann. of Bot.. 

 July, 1906. 



