MUSEUM OF COMPARATIVE ZOOLOGY. 437 



in front. In the Patellidce, the most specialized group of all, containing the 

 largest and most specialized species of Docoglosas, we have the tactile muzzle 

 margin and its prolongations gone, the surface is simply papillose without 

 special margin; the eyes are well developed in the larger species; there are 

 traces of an epipodium, and the hereditary ctenidium, which still lingered 

 with Scurria, has entirely vanished. The circular cordon performs all its 

 functions. The teeth are more fully specialized, all divisions of the radula 

 being represented in the series. Among the Rhiphidoglossa we have an 

 Acmaeid gill in Scutellina, a totally different form of gill in Addisonia, and 

 a pair of symmetrical ctenidia in Emarginula, etc. ; yet no one can question 

 that these animals all belong to the same great group or order. 



In the Petrophilous pulmonates, we have Siphonaria with a gill and lung, 

 Gadinia with only a lung, and Onchidium with a lung and dorsal, possibly 

 branchial, proliferations. In no one of them is there a typical ctenidium 

 developed. These examples illustrate my reasons for considering the breath- 

 ing organs very mutable structures, and unsuited for use in classification of 

 the higher groups, or as diagnostic characters for important subdivisions of the 

 systems. Dr. Pelseneer criticises Dr. Fischer for using the number of the gill 

 ranks or series as a basis for classification. He elsewhere criticises the writer 

 for upholding the view that these characters are not well adapted for such a 

 purpose. Lastly he adopts a diagrammatic classification for the different forms 

 of ctenidia, with the implication that, presented in the manner he adopts, the 

 ctenidia may be used for classification. It seems to the writer, however, 

 that Dr. Pelseneer has really, though unconsciously, adopted the principle 

 upon which Dr. Fischer based his prior classification, merely adding to and 

 correcting the latter in details. Dr. Pelseneer claims that classification by 

 ctenidia is possible if we take the structure and not the number of the lamellae. 

 He then proceeds to show how the number may be more correctly counted 

 than formerly, and having done this uses it as a basis for classification. This 

 is quite proper from his standpoint, but proves that his advance beyond his 

 forerunners is less than he supposes. 



In 1886, I called attention to the manner in which the stages of development 

 of the breathing organs in Cuspidaria, Cetoconcha, Verticordia, and Lyonsiella 

 represent the successive stages of development of the typical ctenidium (Blake 

 Report, Part I., pp. 280, 281, 1886); not by this intending to convey the 

 idea that I believed these groups to be derivative, one from another, but only 

 that they represent the steps by which development in a related series would 

 proceed. I believe this view is correct, although, as will be seen later, I regard 

 the gill in Lyonsiella as of the ordinary type, while the lamellae of Cetoconcha 

 and Poromya are a totally new development, partly consequent on the pre- 

 vious degeneration and obliteration of normal ctenidia in the ancestors of these 

 genera. On the other hand, Dr. Pelseneer would regard the lamellae in these 

 forms as degenerated remnants of the normal typical gill, for which in 

 these observations, I have reserved the name ctenidium. In another place I 

 have shown the anatomical grounds for the belief that this construction of the 



