26o GENERA INCERTAE SEDIS [CH. 



Salvinia Adan. and Azolla Lam. 



One of the outstanding problems in the Natural Classification of the Fili- 

 cales is found in the Hydropterids. A reasonable probability of a relation 

 of the Marsiliaceae to the Schizaeaceae has been stated in Vol. II, Chapter 

 XXIII. But the Salviniaceae, with their two genera Salvinia and Azolla, 

 still remain as ''incertae sedis." Few organisms have ever been subjected to 

 so exact an analysis as these, both as regards structure and development. 

 There can be no doubt of their Fern-nature, notwithstanding their pronounced 

 heterospory. The difficulty in placing them phyletically arises from the fact 

 that, while they have preserved certain features distinctly Filical, these are 

 standardized for Leptosporangiate Ferns at large rather than distinctive. 

 Moreover, their vegetative system is so reduced in accordance with their 

 floating habit, that there is little left in it beyond the initial segmentation upon 

 which exact comparison can be founded. We have learned that comparison 

 based on details of segmentation by itself is liable to pierce the hand of him 

 who uses it, unless there is other evidence to support it. 



There is no need to recapitulate here the details which are so well ren- 

 dered in the original memoirs, and have found their way into all textbooks^ 

 The general facts are indeed fully known. Synonymy, so suggestive else- 

 where of possible affinity, does not help us here. Azolla, with its 5 species, 

 still retains its old generic name introduced by Lamarck in 1783; and 

 Salvinia, with its 13 species, that of Adanson in 1763. It is the soral 

 characters that offer the best ground for comparison. \x\ Azolla filiadoidesNow 

 Goebel finds the female sorus constructed as in Fig. 754 (i) {Organographie, 

 II, 2, 1918, p. 1 132), with a conical receptacle {P) from the base of which 

 rises the enveloping indusium (^Id) composed of two cell-layers, with a distal 

 pore. The receptacle, with its vascular supply, bears distally a single mega- 

 sporangium {Ma), while below, in gradate sequence, are microsporangia which 

 are arrested at an early stage, but show the usual segmentation. On the 

 other hand the male sori may bear an abortive megasporangium distall}-. 

 There appears to be some inconstancy in the occurrence of these in indi- 

 vidual sori, but Von Goebel finds both states in A. filiculoides {I.e. p. 11 32, 

 footnote). He interprets the facts as showing a reduction in number of the 

 megasporangia as compared with that of the microsporangia, and a separa- 

 tion of the originally bisexual sori into those that are male or female. The 

 separation appears to be complete in Salvinia: this genus is peculiar among 

 Ferns in the branched pedicels of its microsporangia (compare Luerssen, 

 I.e. Fig. 186). 



1 For references to the literature see Luerssen, Rab. Krypt. Fl. in, 1889, p. 595. Diels, E. and P. 

 I, 4, 1902, p. 381. Campbell, Mosses and Ferns, 3rd edn. 1918. Strasburger, Tex/book, Engl. Edn. 

 1921, p. 381. ■ 



