482 MONTGOMERY. [Vol. XI. 



teans have developed directly from marine species, though 

 there is no substantial proof for this view. 



Now it should be an aim of investigators who are comparing 

 marine with freshwater and land faunas, to determine from 

 which marine species the others have been derived, and what 

 structural modifications have been evolved by the change of 

 environment. Such an investigation could be best pursued at 

 a locality where a river empties into the ocean ; and by deter- 

 mining what species occur on the neighboring sea-coast, at the 

 mouth of the river, and further up the river, it would be possi- 

 ble to deduce the different degrees of adaptation to the new 

 environment, and the resulting stages of structural modifica- 

 tion. In the case of the nemerteans, such a study could be 

 carried on in Germany at the river Spree, or here in the United 

 States of America in Delaware Bay and the Schuylkill river, — 

 since in both localities we find marine species at the mouth, 

 and freshwater species further up the river. As to structural 

 differences caused by the change of environment, the fact occurs 

 to me that, while the nearest marine relatives of the freshwater 

 forms (with the exclusion of Nemertes polyliopla, Schmarda) are 

 forms with four eyes (genus Tetrastemma), the majority if not 

 all the freshwater forms possess a greater number of eyes (from 

 four to eight). Thus the change from salt to freshwater would 

 seem to necessitate a larger number of eyes, due perhaps to 

 the different conditions of light; and it would be interesting to 

 determine whether the same law holds good in allied groups, 

 — as the Tiirbellaria. There is a fact of importance in this 

 connection, namely, that hand in hand with an increase in the 

 number of eyes of the freshwater nemerteans, takes place an 

 increase also in the variability of their number. While among 

 about a hundred individuals of the marine Tetrastemma ver- 

 miculum which I examined alive at Newport and Woods Holl,^ 

 the number of eyes was invariably four; and among about fifty 

 individuals of Tet^^astemma candidiim studied at Woods Holl, 

 in only one case I found five eyes; in the freshwater Stichos- 

 temma eilhardi, on the contrary, I found no constancy in their 



^ I wish to express my thanks to the U. S. Fish Commission, for the use of a 

 room at their laboratory at Woods IIoll, where this article was written. 



