No. 2.] FRESHWATER AND LAND NEMERTEANS. 48 1 



themselves gradually to shallower water, — occurring first 

 below tides, then between tides, and finally above high tide 

 level ; or they might have accustomed themselves to salty 

 marshes, which in certain periods of the year become dry, so 

 that the worms would be left on dry land for shorter or longer 

 periods, and then adapt themselves to land which is never 

 covered by water. But though this might possibly be the 

 source of the land nemerteans found near a sea-coast, as the two 

 species occurring respectively on the Pelew Islands and on the 

 Bermudas, we must rather expect that those forms occurring at 

 a greater distance from salt water, and in the neighborhood of 

 rivers and lakes, have descended from freshwater species. 

 That is, individuals of a freshwater species, becoming by 

 degrees accustomed to shallow water and marshes, which are 

 subject to periods of drought, would gradually accommodate 

 themselves to dryer places, until they would be able to live 

 on land which is but slightly moist. That the freshwater 

 origin is the more probable for most if not all land nemerteans, 

 is based on three reasons : (i) if they were descended directly 

 from marine species, we would expect the individuals to 

 increase in number as we approach the sea-coast, which, how- 

 ever, does not seem to be the case ; (2) the land nemerteans 

 (as is also the case in the Tttrbellarid) are numerically fewer 

 than the freshwater forms, while if they had developed directly 

 from marine species, they should be at least as numerous if not 

 more abundant than the freshwater forms. And thirdly, we 

 should expect that the land nemerteans are more closely allied 

 to freshwater than to marine forms, because land forms are 

 derivatives of freshwater forms in all those groups of the 

 Invertebrates, which are of marine origin. Thus the closest 

 relatives of the land Turbellaria {Triclads and Rhabdocoeles^) 

 are freshwater species ; the same holds good for the land 

 Annelids {Oligochaetd), and for the land pulmonate Mollusca. 

 Accordingly I feel justified in concluding, that as the fresh- 

 water nemerteans are derived from marine species, so the land 

 nemerteans are derived from freshwater forms ; there is, of 

 course, the possibility that a small number of the land nemer- 



^ Only one land Rhabdocoele is known, Prorhynchus sphyrocephalus, v. Graff. 



