On the connective tissues and body cavities of the Nemerteans. 39 



adds that T. clepsinoides is "wahrscheinlich dieselbe Art", as those 

 insufficiently described forms from Plön, Hamburg, Graz, Cherwell, 

 Canal Saint-Martin, Dorpat, Tascbkend, and Africa — though Stuhl- 

 mann's descrii)tion of the african species e. g., reads merely "eine 

 4äugige Nemertine" ; Bijrger's assumption is unwarranted, that these 

 freshwater forms, whose structural details are as yet wholly unknown, 

 are of the same species. In fact, judging from the structural diver- 

 sity of those freshwater and land forms, which are at all adequately 

 described, we must acknowledge not only various faunal derivations 

 (of. Montgomery, '95 b), but also a phylogenetic development from 

 dififerent marine species, and probably even genera. 



I accept Burger's ('95a) correction of my ('95a) unfortunate 

 division of the Tetrastemmatidae, based on differences in the position 

 of the mouth-opening; at that time, not having had opportunity to 

 study marine species, I relied on the observations of older authors, 

 and their observations were inaccurate. 



In concluding, I would refer briefly to a few points in the classi- 

 fication of the Metanemertini, adopted by BtîRGER in his Naples 

 monograph. The families Eunemertidae, Ototyphlonemertidae, Tetra- 

 stemmatidae (perhaps), Nectoneniertidae, Pelagoneniertidae, and Malaco- 

 hdellidae, as he defines them, are morphologically well characterized. 

 But his family Prosorhochmidae, comprising the very different genera 

 Prosorhochmus, Prosadenoporus^ Geonemertes is a very heterogeneous 

 group; thus the first is hermaphroditic and viviparous, with 4 eyes, 

 without neurochord cells; the second is hermaphroditic and ovi- 

 parous, wilh 4 eyes, without neurochord cells, and apparently without 

 nephridia (Bürger has not found them, though he supposes them to 

 be present!); the third is oviparous, diöcisch or hermaphroditic, with 

 always a variable number of eyes — in australiensis as many as 40, 

 with or without (?) a nephridial system. There is in fact hardly a 

 good character, which these genera have in common 0- Further, I 

 agree with Bürger in the advisability of limiting Geonemertes to 

 the species palaensis and chalicophora, and founding a new genus for 

 australiensis — leaving to him, of course, the right to name the genus; 



1) I would correct a contradiction in Bürgkb's monograph, where 

 he says (p. 143): ''Prosorhochmus ist wahrscheinlich getrennten Ge- 

 schlechts, indess hat man bislang nur Ç$ aufgefunden"; and (p. 553): 

 ''Die Prosorhochmus-Arten sind wahrscheinlich meistens Zwitter"; the 

 latter statement is, of course, the correct one. 



