194 BROOKLYN BOTANIC GARDEN MEMOIRS 



ture and tetrasporic, 7 were of the "Vepreculae" structure and cysto- 

 carpic, 5 were of the "Vepreculae" structure and antheridial, and i 

 was of the "Vepreculae " structure and apparently sterile. Of 9 plants 

 (no. 6966) found growing together at the mouth of Guanica Harbor, 

 Porto Rico, 6 were of the "Vepreculae" structure, 2 of them being 

 obviously antheridial, and 3 were of the " Brachycladia " structure, 

 2 of them obviously tetrasporic. Of 5 specimens (no. 7468) found 

 near low-water mark on Muertos Island, Porto Rico, 2 were of the 

 "Brachycladia" structure and tetrasporic, and 3 were of the "Vepre- 

 culae" structure, i being cystocarpic, i antheridial, and i apparently 

 sterile. In some cases, a considerable series of specimens, all of one 

 group, has been collected, but in collecting the red algae it often 

 happens, as is well known, that the plants found at one time and place 

 may be either all tetrasporic or all sexual. Without waiting for the 

 results of cultural experiments which might furnish absolutely complete 

 proofs of the suggested genetic continuity, it seems to the writer that 

 the evidence is overwhelming that the so-called species of the Kjell- 

 man's "Vepreculae" section are simply the sexual phases of the species 

 of the " Brachycladia " section. It is of interest to note that B0rgesen, 

 in a recent instalment of his admirable series of papers on "The Marine 

 Algae of the West Indies,"^ relying upon the sectional distinctions 

 proposed by Kjellman, appears to have described and figured the 

 antheridial plant (sect. "Vepreculae") of Galaxaura marginata (Ell. 

 & Sol.) Lamour. as a new species under the name Galaxaura occi- 

 dentalis B0rg., taking the tetrasporic plant (sect. "Brachycladia") 

 to be the true G. marginata.'^ 



When we come to examine the alleged species of some of the 

 other sections of the genus Galaxaura, as monographed by Kjell- 

 man, we find strong evidences of other correlations similar to those 

 already described for the Cameratae-Spissae and Brachycladia- Vepre- 

 culae groups. In Kjellman's section "Rhodura," the peripheral 

 elements of the thallus are so manifestly and predominantly fila- 

 mentous (text-figure 3) that there is little ground for using the term 

 "cortex" in connection with these plants, yet there is commonly a 



•'•2: io()-\iT,. f. 118-123. 1916. 



* The original of the Corallina marginata of Ellis and Solander (Nat. Hist. Zooph. 

 115. pi. 22. /. 6. 1786) was from the Bahama Islands, and, like most of the Ellis 

 and Solander types, it is not certainly known to be now in existence. However, 

 there is, in the herbarium of the Royal Botanic Gardens at Kew, an old fragment, 

 inscribed in the hand of Lamouroux: 



Galaxaura marginata 



Corallina Sol. et Ell. 



Bahame 

 which may or may not represent an authentic bit from the Ellis collection. This is 

 antheridial and has the "Vepreculae" structure. 



