— 140 — 



nearest to the typical incrassata has small utricles bringing it into 

 the monile-simulans-group while the fig. b with broader joints, 

 bringing it nearer to simulans, has large utricles. And another spe- 

 cimen (my collection nr. 1601), in its outer appearance coming very 

 near to monile, had rather large, about 60 ß broad utricles bringing 

 it over into the typica group. Also specimens often occur which 

 both in their outer habit and anatomical characters give them an 

 intermediate place. Thus at the shore in the Lime Tree Bay (Ga- 

 savagarden) I have found specimens which in their outer habit 



Fig. 6. Halimeda tridens (Ellis et Solander) Lamx. 



Two forms from the sea west of Water Island, St. Thomas (my collection nr. 1120). About 2 /j 



were intermediate between H. incrassata, typica and monilis and in 

 the same way the diameter of the peripheral utricles were in one 

 specimen 29 — 54^, in another 40—48^, in a third 35 — 47^. 



By reason of these observations, therefore, I cannot agree with 

 Howe in considering these forms as different species but can only 

 consider them as varieties of the same species. 



