46 R. HOLMAN PECK. 



pairs of memhranous plates placed on a pair on either side 

 the foot. The filamentary structure of the gill is, according 

 to him, secondary and due to a breaking-up of the gill-plate. 

 In support of this view he appeals to an observation of Stepa- 

 noff's,! to the effect that in Cyclas two pairs of small mem- 

 branous ridges precede the proper gill-filaments. Stepanoff's 

 observation seems to me to warrant no such interpretation. 



The chief conclusion to which my observations have 

 tended is the exact reverse of this, and is to the effect that 

 the gill-plates of the Unionidse are a highly modified form 

 derived from a simpler condition, in which the gills consist 

 not of plates but of a series of juxtaposed independent 

 filaments, such as we see in a less modified state in Area and 

 Mytilus. It has already been pointed out by Mr. Lankester^ 

 that these gill-filaments of the Lamellibranchia are the exact 

 homologues of the tentacular filaments of the Polyzoa, and 

 of the filaments on the spiral lophophor of the Brachiopods. 

 To establish the possibility of the development of the more 

 complicated forrns of the Lamellibranch gill from such a 

 series of simple free filaments will be to confirm Mr. Lan- 

 kester's view as to the relationship of the ciliated filaments 

 in the three classes, Polyzoa, Brachiopoda, and Lamelli- 

 branchia; the first two of which are better called Tentaculi- 

 branchia and Spirobranchia (Haeckel) respectively. This 

 will be the main object of the present paper. 



Outline of the Essential Constitution of the La- 

 mellibranch Gill, and the Processes by which 

 it becomes modified. 



The Lamellibranch gill which most nearly retains the 

 primitive filamentary form is that of Area ; the gill of 

 Mytilus is in a condition very nearly as archaic. 



The constitution of the gill of Mytilus was first made 

 clear by the excellent observations of M. de Lacaze Duthers, 

 whose nomenclature will serve as the basis of that adopted 

 in the present paper. 



The gill of Mytilus, like that of all other Lamellibranchs, 

 consists of two pairs of gilUplates placed on each side of the 

 body, and separated by the foot, from the base of which on 

 the right and left side, along a line^ stretching from the 

 mouth towards the anus, each pair of gill-plates arises. The 



^ Stepauoff, 'Arcbiv fiir Naturgesch.,' 18G-4. 



- ' Quarterly Journal of Microsc. Science/ vol. xii, 1873. " Remarks ou 

 Rhabdopleura." 



^ This line, on account of its relationship to the lophopliore of Polyzoa and 

 Brachiopoda, Mr. Lankester proposes to call the "lophophoral line" or 

 " lophophoral ridgc=" 



