470 J. H. ORTON. 



Thus the one feature of Protobranchia regarded by Pelseneer as 

 unique — the occurrence of gill-leaflets — is undoubtedly robbed of its 

 glamour — and cannot now be reasonably regarded as of such great 

 taxonomic value as formerly. Another supposedly unique feature of 

 the Protobranchs, namely, the absence of a subdivision of the mantle 

 cavity (see Sedgwick 15, p. 345), must be abandoned, for Drew has 

 already shown that in Yoldia (11, p. 14) there is a subdivision of 

 the mantle cavity into inhalent and exhalent chambers, as is here 

 described for Nucula, and as will no doubt be found in all the Proto- 

 branchs. With regard to the other primitive or special features of 

 Protobranchs, such as the occurrence of distinct pleural ganglia, a 

 plantar surface to the foot, free communication between the cavities 

 of the gonad, pericardium, and kidneys, and other features — with 

 regard to these, Pelseneer has already shown (12, ^jassi?/i) that they 

 are matched in some adult members of the Pilibranchs with the 

 exception of separate pleural ganglia, which at present are only 

 known in the developmental stages of other forms (as Modiolarca, 

 Dreissensia, etc. See Pelseneer 12, p. 234). It is therefore clear 

 that the Protobranchs cannot now be classified as a group equiva- 

 lent to the Filibranchs. We must therefore be prepared to degrade 

 — or rather elevate — the Protobranchia to a subordinate position in the 

 Filibranchia of Pelseneer, or in the Eleutherorhabda of Eidewood. It 

 is a matter of much interest that Palaeontologists (16, p. 359) should 

 already have classified together the Protobranchia and the re- 

 mainder of the Filibranchs into the order of Prionodesmacea, whose 

 diagnosis is concerned mainly with shell characters. As we may 

 now take for granted that Lamellibranchs have evolved mainly on 

 the principle of perfecting the gill as a feeding organ, it is clear — 

 from the closely similar results attained by Palaeontologists and 

 modern zoologists — that there is a close correlation between shell 

 characters and gill characters. Whether the shell characters are 

 capable of any functional explanation, similar to that of the gill 

 characters, my knowledge of the group does not yet enable me to 

 say. It is probable that such an explanation may now be possible. 



XII. A COMPAEISON OF THE MODE OF FEEDING IN 

 LAMELLIBEANCHS AND CEEPIDULA. 



In Lamellibranchs, as in Crepidula, it has been noted that there is 

 an arrangement whereby an automatic selection of the heavier par- 

 ticles takes place just inside the inhalent chamber. Thus the 



