—8 4 - 



if the aquatic form ever changes into the terrestrial form or fruits" or "if the so- 

 called terrestrial R.fluitans and the aquatic R.fluitans belong to the same species. " 

 Here again, it would be interesting to know just how the two terrestrial plants 

 differed from each other, but no distinguishing morphological characters are 

 given. It is stated, however, that the reticulate spores of the so-called terres- 

 trial R. fluitans measure 75-90 \x in diameter, and this would seem to show 

 that the material had been correctly determined and had not been confused 

 with the closely related R. Huebeneriana or R. Sullivantii, the spores of which 

 are smaller. 



Although Donaghy's observations throw doubt on the present delimitation 

 of R.fluitans, he evidently considers that the so-called aquatic form of the species 

 is something definite. The contrary opinion is clearly expressed by Familler 

 (3, p. 12), and he presents certain morphological distinctions to support hisviews. 

 According to his statements it has now been definitely established that Riccia 

 fluitans consists of at least two or three aquatic forms of various Ricciae. He 

 therefore does not accept " R. fluitans" as a species but ascribes specific rank in- 

 stead to what most writers regard as the fertile terrestrial form of R. fluitans. For 

 this he takes up the old name R. canaliculata Hoffm. (as some of his predecessors 

 have done) and calls the aquatic form of this species forma fluitans (L.) Fam. 

 Under R. Huebeneriana he includes Torka's var. natans, as Riccia Huebeneriana 

 forma natans (Torka) Fam., and he recognizes further an aquatic form of R. pseu- 

 do-Frostii. Since, however, he regards this species as a variety of R. Huebeneriana 

 its aquatic form receives the name Riccia Huebeneriana var. pseudo-Frostii forma 

 natans Fam. He implies that other species of Riccia may likewise have aquatic 

 forms and suggests that a Mexican specimen, collected by W. Schaffner many 

 years ago and determined by A. Braun as Ricciella fluitans, may be the aquatic 

 state of R. Frostii, but he does not give this plant a formal name. 



In separating the aquatic form of R. canaliculata from those of R. Hueben- 

 eriana, he assigns to the first very long air-spaces and to the second shorter 

 and more crowded air spaces; in separating the aquatic form of the var. pseudo- 

 Frostii from that of the ordinary R. Huebeneriana, he assigns to the first marginal 

 air spaces about as long as wide and a clearly defined median band and to the 

 second marginal air spaces about twice as long as wide and an indistinct median 

 band. These differences are shown on pi. 1 and 2; the figures give the impression 

 that the morphological distinctions between the various aquatic forms are based 

 on variable characters and that careful culture studies would be necessary before 

 the taxonomic value of these characters could be accurately estimated. Of 

 course, if it could be proved that an aquatic form with long air spaces gave rise 

 to typical R. canaliculata and that a form with short spaces gave rise to R. Hue- 

 beneriana, the differences in the air-spaces would at once assume a greater im- 

 portance. Unfortunately proof of this character is lacking, and some of Von 

 Gaisberg 's culture experiments, to be mentioned below, have yielded contradictory 

 results. 



Von Gaisberg 's studies on Riccia (4), carried on in Goebel's laboratory 

 at Munich, throw an interesting light on some of the questions here discussed. 



