224 



PHILOSOPHICAL NOTES ON 



of a leaf, as the bracts or scales of the Lycopodium have been 

 called by a sort of promotion from scales to bracts. Nothing can 

 better illustrate the genesis of a one-celled anther than the sorus 

 of this Lycopod. It is, moreover, located in the axilla of a bract 

 (Fig. 76), just as stamens often are located in the axilla of a petal 

 or bract, such as we see in Myrlca, Fig. 77. 



In the genus Salvia the one-celled anther is carried away from 

 its fellow by a widely diverging fork of the stamen, much in the 

 way that the two sori of Gynmogramme leptophylla, Desv., 

 Fig. 78, are on distinct prongs of the venous fork. 



Fig. 77. Bract and stamens of Myrica, 

 Gale. (Le Maout and Decaisne, p. 682). 



Fig. 78. Sori of Gymnogr amine 



lepiophylla, Desv. 

 (Hooker's " Brit. Ferns," pi. 1). 



We may perhaps think too much of the fact of sori appearing 

 on the dorsal aspect of the fern leaf as a great difference between 

 cryptogams and phagnogams, which give off their pollen and 

 ovules from the ventral aspect. But if we consider what the fern 

 leaf really is — a stem with its branchlets united by mesophyl, 

 which,' in other words, means only a fasciation of its branchlets, we 

 shall think little of this difference. 



•Then in Acrostichum (including Olfersia) the sporangia are 

 either on the upper surface (ventral) or on both surfaces. And in 

 A. peltatum the fructification is exactly like that of Dorstenia, on 

 a peltate expanded receptacle, fringed by pinnules. In Olfersia 

 the sporangia cover both surfaces of the midrib (dorsal and 

 ventral), while in Trichomanes the sporangia are situated all 

 round the vein, which projects, beyond the pinna, denuded of 

 mesophyl, in the form of a bristle. 



