422 STUDIES IN FOSSIL BOTANY 



A more general discussion of the affinities of 

 Pteridosperms will follow later ; as regards the 

 Lyginodendreae it is evident that their affinities, on 

 the •Cryptogamic side, are altogether with the Ferns, 

 though doubtless they sprang from that stock long 

 before any of the Fern families now familiar to us had 

 come into being. On the other hand, the near relations 

 between the Lyginodendreae and the Cycadophyta are 

 manifest, as shown, above all, in the structure of the seed. 



The Lyginodendreae have also been named the 

 Sphenopterideae, and there can be little doubt that a 

 large proportion of the Fern-fronds of the Sphenopteris 

 type belonged to this Pteridospermous family. If, 

 however, we may take the Crossotheca type of microT 

 sporangiate organ as characteristic, the family was not 

 limited to Sphenopterid foliage, for various species of 

 Crossotheca are known to have been borne on Pecopterid 

 fronds. 1 As regards seeds referable to Lyginodendreae, 

 it appears certain that the various species of Lagenostoma 

 and some of those in Williamson's genus Conostoma, so 

 similar in all essentials to the seed of Lyginodendron 

 oldhamium, must have belonged to the same group of 

 plants. This probably also holds good for Physostoma, 

 in spite of its unique peculiarities (see above, p. 3g6). 2 



A very remarkable fructification which may prove 

 to have affinity with this family is that described by 



1 Three such species are enumerated by Mr. Kidston, Microsporangia of 

 Pleridospermeae, I.e. p. 432, besides two other Pecopterids, regarded by 

 M. Zeiller as probably belonging to Crossotheca. One of the latter, Pecopteris 

 exigua, is especially interesting, as the structure is known, and bears 

 considerable resemblance to that of Telangium Scotti. See Renault 

 Cours de bot. foss. t. iii. Plate xix. Figs. 13-18. 



2 See Prof. F. W. Oliver " On Physostoma elegans," Ann. of Bot. vol. 

 xxiii. January 1909. 



