72 



GYMNOGRAMMOID FERNS 



[CH. 



it may be doubled. Such conditions show that here the sporangium is far from 

 being standardised. The same holds for its spore-output. This is often stated 



Fig. 630. Ceralopteris Ihalictroides : vertical sections of young sporophylls to show the 

 superficial origin of the sporangia {s). ( x 100.) 



as sixteen, but Benedict found it to be definitely thirty-two for many plants, 

 while Thompson (Jx. p. 389) found the spore-count in his material to be 

 anything between twenty-four and twelve. This again shows that the 

 sporangium of Ceratopteris is not a constant entity. 



Fig. 631. A — .£ = sporangia of Ceratopteris thaUctroides, selected to show the variability 

 of the annulus. y^, C = spores. (After Dr J. McL. Thompson.) 



Comparison of the Prijviitive Gymnogrammoid Ferns 



Llavea has always been grouped with Cryptogramme, while J. Smith's name 

 for it, Ceratodactylis osmimdioides, points the comparison also with Osmunda. 

 Sir W. Hooker {Sp. Fil. Vol. Ii, p. 125) remarks, "This is in every respect 

 a very striking plant, closely allied in its fructifications to Cryptogramme Br. 

 but with a very different habit, in some respects approaching Osmunda." 

 None of the early writers compare it with Plagiogyria: but Diels (E. and P. 

 I, 4, p. 255) places these three genera Plagiogyria, Llavea, and Cryptogravime 

 in close relation, which is probably correct. It has already been pointed out 



