326 THE PEARLY FRESHWATER MUSSELS— SIMPSON. vol. xvm. 



R' 



ber ending in the anal siphon is completely separated from the mantle 

 cavity; a7ial and superanal eavity ioiifed, continuing hacl'ward over the 

 adductor muscle into a superanal chamber. Mantle open or closed into 

 more or less perfect siphons, sometimes united for some distance forward. 

 Embryo a lasidium, composed of three segments, the anterior head-like, 

 the median bearing a single shell, the posterior tail-like. 



It will be seen from the above that the characters of the soft parts 

 are qnite variable, and I have italicized those in both shell and animal 

 which seem to most constantly differ from the same in the TJnionid;p. 

 It is very probable that with a more thorough anatomical knowledge 

 of the Naiades the descriptions will have to be a good deal modified.^ 



The following is a list of the genera I place in this family: 



Mtttela, Scopoli. 

 Chelidonopsis, Ancey. 

 Spofha, Lea. 

 PJeiodoii, Conrad. 

 Brazziva, Bourguignat. 



Glabaris, Gray. 

 Iheringella, Pilsbry. 

 Monocondylwa, d'Orbigny. 

 Fossula, Lea. 

 Mycetopoda, d'Orbigny. 



Although in time past the Naiades or pearly fresh -water mussels 

 have often been placed in a single family, and though even v. Ihering, 

 whose recent classification of the genera is, I believe, a natural one, 

 has placed the two groups, Unionidie and Mutelidie, in one super- 

 family, and notwithstanding the fact that there are a few genera whose 

 position on account of our lack of knowledge is doubtful, yet I think 

 it quite probable that the relationship between these two great groups 

 is not a very close one. 



It is true that the animals themselves do not seem to altogether 

 bear out this assertion. The character of the jn-esence or absence of 

 siphons, on which the families have generally been founded is, as I think 

 I have conclusively shown, utterly variable and worthless. There is 

 usually some distinction in the form and the union or nonunion of the 

 labial i)a]ps, but these characters are not perfectly constant, and even 

 if they always held good, they would be of little importance. Ihering 

 is authority for the statement that in all the South American and Afri- 

 can Mutelidie (and all the genera belong in these two continents) the 

 outer gill-leaves on each side are firmly attached both to the mantle 

 and abdomen, thus completely separating the suprabranchial cavity 

 from that of the mantle back to the anal opening.. This, however, 

 according to that most excellent authority, occasionally occurs in the 

 Unionidic of the northern hemisphere. 



'There will doubtless be found great variation in the matter of union of the mantle 

 and gills in many other Pelecypods. Jackson observed in some specimens of Perna 

 . epliippinm that the two pairs of gills were separated from one another throughout 

 their length, whereas in other specimens the two median gills were connected by c<m- 

 crescence at their dorsal border, thus uniting the two pairs as in Ostrea. The degree 

 of union varied indifferent specimens, the gills being united for their whole extent, 

 or only posteriorly. (Phylogeuy of the Pelecypoda. Mem. Bost. Soc. Nat. Hist., 

 IV, No. VIII, p. 326, 1890.) 



