48 THE DISPERSAL OF SHELLS. 



should be noted, from the localities in which they are 

 deposited or the nature of the objects to which they are 

 usually attached, are not at all likely to be dispersed by 

 animals, and it will be remembered that some molluscs 

 are viviparous. The fry, probably, are transported more 

 frequently than ova. Those UnionidcB whose larvae are 

 parasitic for a time upon fishes are eminently liable to 

 wide dispersal, at least, through waters in any way com- 

 municating with their habitats. The larvae of A nodonta 

 swim by the flapping of the valves of their shells, trailing 

 at the same time a long byssal filament, and weaving 

 themselves together in masses which spread out like a net. 

 Little fishes, stirring up the masses, carry many larvae 

 away, the creatures managing to effect a hold — chiefly to 

 the fins — by means of the toothed edges of their shells. 

 They then become encysted, and are nourished, it is 

 stated, by the exudations of the fish. Ultimately, on 

 the splitting of the cyst, they drop out and fall to the 

 bottom. In aquaria, Anodontce which have retained 

 their brood in long quiescence, have been observed to 

 let it out immediately on the introduction of fishes. 

 The larvae of Unio^ differing in certain respects, are said 

 to be inhaled by fishes, and are usually found attached 

 to the gills.' Whether these fry are likely to be carried 



^ E. Ray Lankester, art. Mollusca, " Encyclopgedia Britannica," 

 ed. 9, xvi. (1883), 694 ; and see also Dr. Schierholtz's memoir as to 

 the development of the Uniomda:^ Imperial Academy of Science, 

 Vienna, 1888, my knowledge of which is derived solely from a 

 leaflet printed by Mr. Standen, in explanation of specimens illus- 

 trating the life history oi Anodonfa, shown by him at one of the 

 Manchester Microscopical Society's soirees. 



