880 PROF. G. 0. BOURNE ON THE [Nov. 17, 



that they have probably descended from them. But this con- 

 clusion cannot be sustained. In some important respects the 

 JSTeritidae are more primitive than any of the flhipidoglossa . 

 The ccelom, in particular, extending as it does across the whole 

 width of the body, retains features which may be called embryonic 

 when we compare it with von Erlanger's and Miss Drummond's 

 account of the development of Pcdudina, but must surely be 

 regarded as primitive when we consider the probable phylogeny 

 of the Gastropoda. No such extensive ccelomic space has been 

 described in any other gastropod, and to find a parallel to it we 

 must refer, as Thiele has done, to the Cephalopoda. When a 

 feature is shared by representatives of two orders now widely 

 separate, and is also shown by embryological evidence and by 

 a pnori reasoning to be primitive, there are very good grounds 

 for regarding it as ancestral. In other Rhipidoglossa the coelom 

 is reduced to a pericardial sac suirounding the heart. This 

 pericardial sac, as embryology teaches us, is the reduced repre- 

 sentative of a primitively much more extensive space. This more 

 extensive space is preserved in the Neritidfe, and the conclusion 

 is that they have inherited it from ancestors more generalized in 

 this respect than the remainder of the existing Rhipidoglossa. 

 This ancestor must have been older even than Pleuroto^naria, for 

 the ccelom is reduced to a pericai-dial sac in this genus. 



The same conclusion is reached by a considei-ation of the 

 excretory organs. In Pleurotomariida?, Haliotidaj, Trochidse, 

 Turbonidse, and Fissurellidse the post-torsional left kidney is 

 reduced to a small sac, and in the first four families this 

 " papillary sac," as it is called, has undergone modification. It 

 no longer serves for the elimination of waste matters from the 

 blood, but is phagocytic. In the Neritidse the left post-torsional 

 kidney is large and persists as the functional excretory organ ; it 

 is the right kidney that has changed its function and undergone 

 reduction. It cannot be doubted that the ancesti-al Gastropod 

 possessed paired functional kidneys as do the Chitonidpe and 

 among the Fissurellida" Cemoria {fide Haller). The obvious 

 inference is that the families in which the left kidney is modified 

 to form a papillary sac are to that extent modified, and that the 

 Neritidpe are descended from an ancestor in which this modifica- 

 tion had not yet taken place. The persistence of the left kidney 

 in Neritidpe, therefore, is to be i-egarded as an ancestral rather 

 than as a specialized character, and as evidence that this family 

 cannot have been descended from Trochidse or Turbonida?, in which 

 specialization has taken a difi'erent direction. 



These arguments indicate that the Neritidse are descendants of 

 a very primitive stock (a conclusion sufficiently supported by 

 Paljeontology), from which the remainder of the Rhipidoglossa 

 and probably other groups of Gastropoda were also derived. 



If these conclusions are accepted, the question of the relation- 

 ship of the aSTeritidse to the Tajnioglossa is simplified. In this 

 case we have to consider whether the special characters of the 



