REPORT ON THE MOLLUSCA. 25 



The structure of this branchial partition is entirely muscular ; at the point wliere 

 the apertures occur it is somewhat less thick. The epithelium of the ventral surface of 

 the partition is continued on the lips of the apertures ; it is ciliated there, but has no 

 special modification. 



As in Poromya and Silenia, the muscular partition has, on its outer borders, 

 delicate muscular bundles which are attached to the shell (PI. IV. figs. 1, 3, 4, r). 



Summary on the Geneea Poromya, " Silenia," and Cuspidaria. 



A. Classification. — The genus " Silenia " is placed by Smith ^ among the Ana- 

 tinidse, and by Fischer,^ doubtfully, among the Lyonsiidje. But the structure of the 

 gills in these two families is absolutely unlike that observed in Sile7iia. 



Among recent authors Jefi"reys,' Sars,* and Smith' range Poromya and Cuspidaria 

 [ = Nesera) among the Corbulidae, and Fischer* classes Poromya in the Anatinidse. 

 But here, again, the gill-structure of Poromya and Cuspidaria is quite different from 

 that of the Corbulidae and Anatinidse. 



What strikes us first in the three genera cited, is the presence of a muscular parti- 

 tion extending from one adductor to the other, and taken by previous authors for the 

 body-wall. We have shown that this partition is formed by the gills, which have lost 

 the structure habitual to these organs in the Pelecypoda, and, consequently, form an 

 arrangement quite unique among them. 



Among the Anatinacea, where the three genera in cjuestion are ranged, they form 

 a sub-group quite different from the other Pelecypoda beside which they have been 

 placed. These others (Anatinacea, s. str.) are true " Lamellibranchia ; " that is to say, 

 they have gills of the typical, normal structure. I propose, therefore, to designate this 

 abnormal sub-group Septibranchia. 



The structure of this group remained unknown, in consequence of the habit the 

 majority of conchologists have of not troubling themselves about the soft portions. In 

 the present instance this has resulted in the classifying of the forms in question as 

 described above ; a fact which shows that we need expect little assistance from the shell 

 in determining the systematic position of a mollusc. 



I have been able to study the structure of each of the three genera, and I ha^•c 

 proved that Poromya and "Silenia" are more nearly allied to each other than to 

 Cuspidaria. In the former the apertures in the partitions are in groups ; the 



' Zool. Chall. Exp., part xxxv. p. 75. ^ Manuel de Conchyliologie, p. 1164. 



■' British Coucbology, vol. iii. pp. 45, 47. * Mollusca rcgionis arcticas Norvegiae, pp. 8o, 96. 



'^ Zool. Chall. Exp., part xxxv. pp. 35, 54. « Manuel de Conchyliologie, p. 1172. 



(zooL. CHALL. EXP. — PART Lxxiv. — 1888.) Eece 4 



