HOWE: DIMORPHISM IN GALAXAURA 197 



be surmised that Galaxaura oblongata (Ell. & Sol.) Lamour.'^ has its 

 tetrasporic condition in G. comans Kjellm. And just as the line of 

 demarkation between Galaxaura oblongata and G. cylindrica seems a 

 little uncertain and arbitrary, so also is the line of separation between 

 G. comans and G. lapidescens. Where there is so much difference in 

 habit as there is between the "Rhodura" forms on the one hand and 

 the "Microthoe" and " Eugalaxaura " forms on the other, there is 

 manifestly more need for a cultural demonstration of their correlation 

 as alternating generations than there is in the case of the Cameratae- 

 Spissae and Brachycladia-Vepreculae groups, where the two 

 phases have the same outward appearance. But while experimental 

 demonstration or further observations in the field may be desirable 

 for a precise correlation of the " Rhodura" forms, the existing evidence 

 that these "Rhodura" forms represent tetrasporic phases of "Micro- 

 thoe" and "Eugalaxaura" forms seems convincing. 



EXPLANATION OF PLATES III AND IV 



Plate III 



Photograph of a formaUn-preserv'ed antheridial specimen of Galaxaura marginala 

 (Ell. & Soland.) Lamour., representing Kjellman's section " Vepreculae." Speci- 

 men from San Juan, Porto Rico {Howe 2J04); natural size. 



Plate IV 



Fig. I. Photograph of a formalin-preserved tetrasporic specimen of Galaxaura 

 marginala (Ell. & Soland.) Lamour., representing the section " Brachycladia." 

 Specimen from San Juan, Porto Rico {Howe 2J04); natural size. 



Fig. 2. Photograph of a formalin-preserved specimen representing a form of 

 Galaxaura flagelHformis Kjellm. and belonging in Kjellman's section "Rhodura." 

 Specimen from Santurce, Porto Rico {Howe 1859a); natural size. 



Fig. 3. Photograph of a formalin-preserved specimen of Galaxaura squalida 

 Kjellm. Specimen from Santurce, Porto Rico {Howe iS^gb); natural size. 



^ Galaxaura fragilis of Kjellman and of B0rgesen; not Dicholomaria fragilis 

 Lamarck, the type specimen of which in herb. Mus. Paris, appears to have the 

 structure of the "Spissae" group. 



