The President's Address. By Duleinficld H. Scott. 145 



Weber and Sterzel have shown that these petioles, in their turn, 

 were borne on the stems known as Medullosd. This is well seen 

 in the English species {Mcdtdlosa anglica) where the leaf-bases 

 attached to the stem show the typical structure of a Mycloxylon. 



The organisation of the stem is peculiar, for it contains several 

 •distinct vascular cylinders, thus resembling the stem of the more 

 complex Ferns, but each of these cylinders grew in thickness on 

 its own account, by means of a special cambial layer — a combina- 

 tion of characters not known in any living plant. While the 

 primary structure of the stem was Fern -like, the secondary tissues 

 resembled those of Cycads, as was also the case with the structure 

 •of the petiole ; the form and venation of the leaves, however, take 

 us back once more to the Ferns. In the case of the English 

 Jfcdullosa the foliage was that of an Alcthoptcris. Owing to these 

 indications of affinity in two directions, the family Medullosere, or 

 Neuropteridea3 (for the names may be treated as synonymous), have 

 been classed of late years in the intermediate group Cycadofilices.* 

 Thus Stur's exclusion of these plants from the true Ferns has been 

 justified by anatomical evidence. 



Suggestions as to the probable fructification were made in 1898 

 and 1900 by Hemingway and Wild, the former relying on the 

 association of the fronds with certain unassigned seeds, while the 

 latter called attention to some points of structural similarity 

 between one of these seeds, Trigonocarpon, and the petioles of 

 Mcdullosa, with which it constantly occurs in association. The 

 •connection, however, of these seeds with the Neuropteridese re- 

 mained doubtful, though very probable, until strengthened by 

 Mr. Kidston's striking discovery, in 1903, of a large seed in actual 

 continuity with the pinnules of a well-known species of Ncuroptcris, 

 N. lictcrophylla, the most impressive example of a " seed-bearing 

 Fern " which has yet been brought to light f (pi. II. fig. 4). From 

 the nature of the preservation it has not been possible to examine 

 the structure of the seed of Ncuroptcris lictcrophylla, but that of 

 Trigonocarpon is well known, at least in its main features. The 

 body of the seed is oval, attaining an inch in length ; it bears 

 twelve longitudinal ridges, of which three are more pronounced 

 than the rest, and have thus given the genus its name.J The 

 ridges belong to the hard fibrous inner layer of the seed-coat ; this 

 was enveloped by a fleshy outer coat, of considerable thickness, 

 bounded by a well-marked epidermis. The fleshy outer testa is 

 only well preserved in the best of the petrified specimens (see 

 pi. III. figs. 5 and 6). Favourable specimens further show that the 



* Scott, " Studies in Fossil Botany," 1900, p. 374. 



t Kidston, " The Fructification of Neuropteris heterophylla" Phil. Trans. R.S. (B), 

 exevii. (1904) p. 1. 



X The ordinary three-cornered specimens, however, are merely structureless casts 

 of the inside of the seed. 



