THE ANCIENT FERNS 



of South Africa and Australia. Leptopteris, which resembles Todea in many ways but 

 differs from it by the fact that the leaf texture is 'filmy' as in the Hymenophyllaceae, is 

 confined to New Guinea, Polynesia and New Zealand and contains about seven species. 

 This very modest array of living forms is in striking contrast to the wealth of the fossil 

 record. The most important information regarding this is still to be found in the classic 

 researches of Kidston and Gwynne-Vaughan (1907-14), who described the structure of 

 all anatomically preserved specimens known at that time, and very few have been added 

 since. Stem and petiole structures, so like those of the living Osmunda that they could be 

 thought of as cogeneric with it, have been found in many countries and at many geo- 

 logical levels, as the following list of'Osmundites' will show: 



Table of species q/" Osmundites with anatomically preserved structure known to Kidston and 



Gwynne-Vaughan^ 1907-14 



Horizon Species 



Tertiary of Paraguay Osmundites Carnieri 



Tertiary of Spitzbergen 0. Spetsbergensis 



Miocene of Hungary 0. Schemnitzensis 



Eocene of Britain 0. Dowkeri 



Lower Cretaceous of British Columbia 0. Skidegatensis 



Upper Jurassic of South Africa 0. Kolbei 



Jurassic of New Zealand 0. Dunlopi 



,, „ O. Gibbiana 



At a still earlier level, other stems and petioles have been found, so like those of 

 Osmunda and Osmundites that they must certainly be referred to the same group though 

 not to the same genus, since they differ from all the living and late fossil genera not only 

 in being larger but also in being in certain respects more complex. In the Permian of 

 Russia alone no less than four distinct genera have been described {^alesskya, Thamnopteris, 

 Anemorrhoea and Bathypteris) . Further back still the resemblance of Osmundaceous 

 sporangia to certain unattached fruit bodies of Carboniferous age may perhaps indicate 

 an ancestral type in the Coal Measures. 



A direct record of this kind is something unexampled elsewhere in the ferns, and it is, 

 indeed, only to be compared with Equisetum among living vascular plants. This is the 

 reason for the special importance of the group in the present context, for, perhaps even 

 more than in Equisetum, we know that it, and perhaps even certain species in it, have 

 existed unchanged for very long periods of time, although its most important evolutionary 

 outburst was at the beginning. 



All the species available in cultivation in the botanic gardens of Europe have been 

 examined, and, in addition, I have received one specimen of Osmunda javanica direct 

 from Malay, though this will not be separately described since it differed in no way from 

 the form of the species available at Kew. It will be convenient to deal with the genera 

 in the order listed above, and the species in the genera in the order of their geographical 

 distribution. 



All the available species of Osmunda are represented in Fig. 277 except for 0. regalis, 

 which can be seen on reference back to Fig. 22 a, p. 36. The chromosomes of the two 

 North American species, 0. cinnamomea L. and 0. Claytoniana L. ( = 0. interrupta) are 

 shown in Fig. 277 a and c. Two species or forms closely resembling 0. regalis but smaller, 



276 



