462 MEMOIRS OF THE CARNEGIE MUSEUM. 



One of these characters has been noticed long ago, and has been discussed in 

 detail by Von Ihering. This is the tendency manifested in certain forms to close 

 the branchial opening in front by a connection of the mantle-margins, which is 

 a character entirelj^ wanting in the Unionida;, but it is found occasionally in both 

 subfamilies of the Mutdidm (Hyriimv and MuIcIukv). But numerous members 

 of this famil}^ do not have this character, and thus it is evident that the absence 

 of this mantle-connection indicates a primitive condition, while its presence is a 

 more advanced stage, expressing the tendency to transform the branchial opening 

 into a closed tube (siphon). 



According to Von Ihering (1898), the mantle-connection closing the branchial 

 opening anteriorly is always present in Castalia. I have been able to confirm 

 this only partially. Of typical species of Castalia I possess the soft parts of six 

 specimens of C. acuticosla: they show the branchial opening closed with one ex- 

 ception, where it is open. In C. undosa, a somewhat aberrant type. Von Ihering 

 describes the branchial as normality closed, but he says that in four out of twenty- 

 one specimens examined the mantle-connection was missing.^ Of this species I 

 have five specimens with soft parts, of which three have the branchial closed, while 

 it is open in two of them. In one of the latter, however, a small one, it may be 

 torn. It must be adrnitted that also in other cases this connection of the mantle 

 may have been torn in life or in preservation.** And thus it might be that normally 

 this mantle-connection is present in Castalia. 



Castalina, according to Von Ihering, varies in the development of this character 

 in the different species and individuals. He has investigated two males of C. 

 nehrinfji, and one had the branchial closed, the other open. In eight females ex- 

 amined by myself, I found the branchial open. In C. martensi Von Ihering found 

 the branchial opening in most cases closed. I have no soft parts of this species. 

 But in two specimens of C. psammoica, which I possess, one has this opening dis- 

 tinctly closed in front, in the other this is not the case. 



Nowhere else within this subfamily has this character been found. Castaliella 

 and Callonaia may possess it, but their anatomy is unknown . Of Hyria and Prisodon 



'In V(in Ihcring's dcscrii)f.ii)ii (1S93, ii. 86) there is a very singular mistake. He speaks of the 

 mantle-c(Hinection " Hniccke " between (he two " siphons," while he undoubtedly means the con- 

 nection in front of the branchial " siphon." Tliis is the one wliich is sometimes missing. The absence 

 of the connection between the anal and branchial openings would be something unheard of; in fact, 

 has never been observed by me in any member of the family. 



" We must recall that similar variations or mutilations with regard to tiie mantle-connection which 

 sejjarates anal and supra-anal openings in the Uniunidw arc well established, for instance in tlie genus 

 Fusconaia. 



