KUNGL. SV. VET. AKADEMIENS HANDLINGAR. BAND 57. NIO I. 39 



to regard them as primitive. The geological record available at present indicates that 

 they existed before the fern-type with large fronds, as exemplified by the Upper Devonian 

 Archaeopteris. The line Psilophyton princeps — Arthrostigma — Cyclostigma — undoubted 

 Palaeozoic Lycopodiales does not seem to present any innate improbability, although 

 there is a complete lack of evidence as regards the spore-producing members and very 

 little is known of the anatomical strncture. On the other side there is Psilophyton Gold- 

 schmidtii, which is probably closely related to Psilophyton princeps; and this form, as set 

 forth above. would seem to furnish us with an intermediate stage required by Lignier's 

 hypothesis. From this point of view, the whole pteridophytic stock would be mono- 

 phyletic, the Lycopsida and the Pteropsida being derived from a common form already 

 vascular. It would not thus be necessary to assume a parallel evolution of a similar 

 vascular system along two different lines. The leaves, on the other hand, would be mor- 

 phologically diphyletic. If the microphyllous habit is regarded as primitive — which 

 does not necessarily mean an acceptance of the antithetic doctrine of alternation — it would 

 not be necessary to derive the large fronds of the Filicales from the leaves of the Lycop- 

 sida through a process of progressive development. In the Filicales only the paleae or 

 other emergences on the rachis would be homologous with the leaves of the Lycopsida. 

 Such emergences, in the shape of hairs or spines, are strikingly common in Palaeozoic 

 fronds: a study of the illustrations given by Stur (1875 — 77, pls. 26, 35) of large fronds 

 from the Lower Carboniferous will show forms which have the rachis and its branches 

 covered with hairs or spines similar to those of Psilophyton princeps. Finally attention 

 may be called to the coincidence that the circinate vernation of the fern-fronds is pa- 

 rallelled in the branches of Psilophyton princeps. 



Leaving these speculations it may be stated that the best evidence of the existence 

 of the beginnings of the filicinean phylum in the Older Devonian flora seems to be offered 

 as yet by Dawsonites arcuatus. These supposed fructifications show no relation to the 

 microphyllous series, although it is possible that they may have been borne on lateral 

 branches of plants of the Psilophyton Goldschmidtii-tyipe. It is possible, as suggested 

 by Prof. Lignier, that real fern-fronds may be derived from forms similar to Dmv- 

 sonites arcuatus, through sterilization and cladodification of the ultimate sporangium- 

 bearing branches. At any råte, it is certain that the sporangia of Dawsonites recall those 

 of certain Upper Devonian and Carboniferous ferns generally considered as primitive, 

 as for instance Dimeripteris, and perhaps Stauropteris, according to Lignier (1908 — 1911, 

 p. 51). The fertile fronds of Archaeopteris may further be compared with our form. The 

 chief points of resemblance between the fertile fronds of certain Primofilices and Daw- 

 sonites arcuatus are the large size of the sporangia and their apical position on branches 

 of special fronds or pinnae without developed laminae. Among the fronds of the Lower 

 Carboniferous and the Upper Devonian, the common occurrence of »modified» fronds 

 bearing sporangia but no flattened pinnules is very striking. So, on the whole, is the rela- 

 tive abundance of fertile specimens of Archaeopteris. In the Lower Devonian, finally, 

 we find frond-like structures bearing sporangia, but no fronds with developed laminae. 

 One can hardly escape the conclusion that the »modified» fertile fronds may represent 

 the primitive state in this case and that the flattened pinnules are a låter development 



