GILLS OF LAMELLIBRANCH MOLLUSCA. 51 



filaments — becomes bridged over, and interlamellar junctions 

 of various kinds effected. In Mytilus these consist of a few 

 outgrowths from the internal or sub-lamellar surfaces of tlie 

 filaments which are confined to a few spots, (see Plate IV, 

 figs. 8, 9, 10), and form as M. de Lacaze Duthiers terms 

 them, " veritables baguettes transversales assez longues.^' In 

 other genera the ciliary junctions of Mytilus and Area are 

 replaced by fibrous and vascular interfilamentar concre- 

 scences, which have the form of transverse horizontal bands 

 running antero-posteriorly at right angles to the filaments 

 (see Plates V and VI). 



When concrescence thus replaces the simple ciliary inter- 

 filamentar junctions of the more archaic type, the few 

 tubular sub-filamentar outgrowths which by concrescence 

 produce interlamellar junction in Mytilus are re|)laced by 

 an exuberant sub-filamentar outgrowth, as in Dreissena and 

 Anodon. This important and abundant outgrowth from the 

 back of the filaments has an almost unlimited capacity for 

 concrescence. It gives rise to long vertical interlamellar 

 junctions (parallel to the filaments) disposed at regular inters 

 vals, and it obliterates that part of the inter-filamentar 

 spaces left free by the narrow horizontal bands which join 

 filament to filament, leaving only a series of irregular passages 

 opening on the supra-lamellar and sub-lamellar surfaces by 

 small inter-filamentar stomata. The loose, highly plastic 

 tissue which thus grows out from the back of the filaments 

 consists of lacunar tissue {vide ante, p. 45) covered by an epi- 

 thelium. Channels and even definite vessels become excavated 

 in it and the primitive hlood-camjing function of the hollow 

 gill-filaments is superseded hy the circtdation which arises in 

 the connected mass of suh-flamentar tissue. AVith this loss 

 of function the filaments loose more and more their primitive 

 tubular character and become merely a supporting skeleton 

 for the exuberant sub-filamentar outgrowth. In Anodon, 

 where this condition is reached, the primitive gill-filaments 

 no longer serve even as accessory blood-vessels. Tiiis asser- 

 tion is contrary to the statements of Posner as to structure 

 and the views of Langer, but it will be justified in treating 

 more minutely of the Anodon-gill. 



Another class of modifications to which the Lamellibranch 

 gill-plate is subject consists in the destruction of the per- 

 fect parallelism of its two lamellse and in the substitution 

 of an undulating or fluted surface, so that the two lamelloe 

 meet along one vertical line, then diverge and then -again 

 meet — the whole interlamellar space of the gill plate being 

 thus thrown into a series of sub-cylindlical cavities,each cavity 



