THE ORIGIN OF GYMNOSPERMS 231 



One of the earliest instances in which there appeared to 

 be some grounds for attributing a seed to a certain frond was 

 discussed by Wild 1 in 1900. This investigator showed that it 

 is exceedingly probable that the well-known seed, Trigono- 

 carpus, really belonged to a Medullosa with the Alethopteris-type 

 of frond, not only on the grounds of constant association, but 

 on the agreement of certain anatomical features common to 

 the seed and the vegetative organs. This attribution still 

 appears to be extremely probable, but definite evidence of the 

 fact has not so far been forthcoming. 



M. Grand'Eury, 2 to whose work reference has been made, 

 has supported this conclusion, and has also shown that equally 

 strong evidence, so far as constant association can be regarded 

 as having weight in this connection, exists for referring the 

 Upper Carboniferous or Permian genera Odontopteris, Linoptens, 

 and Callipteridum to the Pteridosperms. In a more recent paper 

 the same author 3 deals with further data, also derived from 

 association, affecting the frond-genus, Sphcnopteris and the seed, 

 Codonospermum. 



M. Grand'Eury has also called attention to the structural 

 diversity of the Carboniferous seeds which may probably be 

 attributed to the Pteridosperms. The investigation of certain 

 Palaeozoic seeds showing structure, at the hands of Prof. Oliver^ 

 has also led to the same conclusion. 



I have dealt with these recent discoveries at some length here, 

 for this work is of extreme importance as supplying the missing 

 link connecting at least one group of the Gymnosperms with Fern- 

 like ancestors. The Pteridosperms may be regarded as standing, 

 roughly, half-way between the Ferns and the Cycads, though in 

 reality they themselves have crossed the borderland, and are 

 obviously on the Gymnospermic side. Thus the views of 

 Dr. Scott and others, who for some time previously have urged 

 the conclusion that the Cycads, the most primitive of Gymno- 

 sperms still existing, were derived from the Ferns, have received 

 full confirmation. Dr. Scott 5 has pointed out that practically 

 every step in the line of descent of the Cycads can now be 



1 Wild, 1900 ; see also Scott and Maslen, 1906. 



2 Grand'Eury, 1904 1 , 1904 2 . 



3 Ibid. 1906. 



4 Oliver, 1903, 1904 1 , 1904 2 . 



5 Scott, 1903, 1905 1 , 1905 2 , 1906. 



