84 GYMNOGRAMMOID FERNS [CH. 



number of 64: but in P. intramarginalis and andromedifolia it is 32, and in 

 P. hastata it is only 24, and sometimes 16. It will be seen later that a wide 

 variability in spore-counts is frequent in this group. 



A new species of Pellaea from China has lately been shown by Dr C. 

 Christensen to bear characters of special interest, since it raises a doubt to 

 which genus it should be referred. In stipe and sorus it resembles Adiantopsis, 

 the imparipinnate pinnae point to P. andromedifolia ox fiavcscens, their shape 

 to P.flcxuosa. The slightly reflexed continuous or lobed margins imperfectly 

 cover the sori, while the sporangia are large and few, or sometimes solitary. 

 Since it appears to offer characters intermediate between Adianiopsis, Pellaea, 

 Cheilanthes and Notholaena, Dr Christensen has named it Pellaea connectens. 

 As will appear in the later comparisons, Dr Horvat finds in this plant a 

 synthetic type, pointing downwards in the direction of Mohria (Christensen 

 649). 



Prantl {Englers Jahrb. Ill, 1882, p. 416) attempted a systematic arrange- 

 ment of the genus Pellaea based on wide morphological data. He included 

 underit Kaulfuss' genus Cassebeera, which is maintained by Christ as a 

 substantive genus {I.e. p. 154). He also included Doryopteris J. Smith, 

 though this is likewise maintained as a separate genus by Diels {I.e. p. 269), 

 and by Christ also {I.e. p. 162). Taking thus a comprehensive view of the 

 limits of Pellaea, Prantl disposed its constituents in a sequence which in the 

 main was probably natural and phyletic: for it was based upon a com- 

 parison of facts of anatomy and soral development not always regarded by 

 systematists. He placed first those with the sori free {Platyloma, Eu-Pellaea, 

 Cinciitalis), and proceeded to those with occasional reticulation of the blade, 

 and with sori more or less fused, "anastomosantes" {Pteridella, Cassebeera, 

 Dorypteridastruin, Doryopteris, Pteridellastrum). He does not attach great 

 importance to the degree of elongation of the sorus from the margin 

 inwards: but he notes particularly how in Eu-Pellaea and Platyloma, which 

 he places first in his series, the sorus is truly terminal, i.e. that the vein- 

 endings do not extend beyond the sporangia {I.e. p. 405). This grouping 

 appears to be in accord with the view that a distinct receptacle and a 

 position of the sorus at or near to the vein-ending are primitive, while soral. 

 fusion, or a spread of the sorus over the surface of the blade, is a 

 derivative condition. It will be seen later that this view is probably correct. 



Doryopteris J. Smith, 1841 

 This genus, which has a profuse synonymy, includes about 40 species of 

 very wide distribution, but they are chiefly located in the western tropics. 

 They are small rhizomatous Ferns, slightly heterophyllous, with black 

 polished petioles, and with coriaceous blades, often pedate. The sori are 

 sometimes separate, and round: in other species a narrow linear fusion- 



