50 R. HOLMAN PECK. 



originally separate become fused into one continuous piece. 

 This process M. de Lacaze Duthiers has spoken of as " la 

 soudure ;" I shall call it " concrescence," the name used by 

 Haeckel for the similar phenomenon which occurs in the 

 Calcispongiae. 



Posner and those who would advocate the membranous 

 plate as the prototype of the Lamellibranch gill must 

 entirely ignore the remarkable and exceptional process of 

 " concrescence" of which we have excellent examples in the 

 Lamellibranchia in other structures than the gills. It is a 

 matter of observation that the free edges of the mantle unite 

 in the siphonate Lamellibranchs by concrescence to form the 

 siphons and the closed mantle sac. The siphon of the Dibran- 

 chiate Cephalopods is formed in the same way from the middle 

 lobes of the foot. 



An instance of abnormal concrescence in the gills of Anodon 

 was brought under my notice by Mr. Lankester. In this 

 case a torn portion of the inner gill-plate of the left side 

 beyond the posterior edge of the root of the foot had become 

 intimately adherent by concrescence to the inner surface of 

 the inner gill-plate of the right side of the animal. 



There can be no doubt after a survey of the facts, that 

 concrescence plays a most important part — in fact, the essen^ 

 tial part in the modification of the primitive gill-filaments of 

 the Lamellibranchia. 



As fully pointed out by M. de Lacaze Duthiers, the free 

 outer lamella of the outer gill-plate and the free inner lamella 

 of the inner gill-plate become not merely adherent to but 

 vascularly continuous with neighbouring structures by " con- 

 crescence" or as one might term it " self-grafting." 



lu the genera Ostrea, Anodon, Pholas, Lutraria, as deter- 

 mined by M. de Lacaze Duthiers (' Annales dcs Sci. Nat. 

 Ser.,' iv, vol. ii, p. 155), the external lamellae of the external 

 gill-plates are not free as in Mytilus, but are fused to the 

 adjacent surface of the mantle by concrescence. Similarly 

 the inner lamella of the inner gill-plate instead of remaining 

 free as in Mytilus becomes in Anodon and other genera fused 

 for a short space to the foot and w here the foot ceases to divide 

 the inner lamellae of the inner gill-plates of each side from 

 one another, their typically-free edges unite by concrescence, 

 producing the partition betwx^en the cloacal and branchial 

 chambers of the mantle. 



Concrescence is not, however, confined in its operations to 

 the production of adhesions of the gill-filaments to mantle 

 and foot. 13y its operation the interlamellar space — the space 

 between the descending and the ascending portions of the 



