ortmann: south American naiades. 463 



it is positively known that no such mantle-connection exists, and the same is the 

 case in all investigated species of Diplodon. 



Thus it is seen that Caskdina and Custalia (and possi]:)ly also Castaliella and 

 Callonaia) form a group within the subfamily, which is not only sjiecialized in the 

 shape of the shell, but also in the tendency to close the branchial opening an- 

 teriorly. As Castalina forms a transition from Diplodon to Castalia in the shell, 

 so it is transitional also in this latter character, and thus it is perfectly clear that 

 Castalina and Castalia carniot be primitive tj'pes, and otiier peculiar characters, 

 which they possess, probably also should not be regarded as primitive. 



This is esiJecially true of the parallel ridges vertical to the edge of the lateral 

 hinge-teeth (also sometimes present on the pseudocardinals). This character is 

 best developed in Castalia (Plate XXXIX, figs. 8c, 8d), but Castaliella is also de- 

 scribed as showing traces of it, and Callonaia has it. But here again Castalina 

 is transitional toward Diplodon. Castalina nehringi has, according to Von Ihering, 

 hinge-teeth obliquely corrugated, but never with i^arallel ridges vertical to the 

 edge. This is quite correct, as far as concerns the laterals; but some of my 

 specimens show that the furrows of the cardinal teeth occasionally may be nearly 

 parallel, thus producing nearl.y parallel ridges. In Castalina marlensi vertical 

 ridges are present and well-developed, at least in the anterior section of the lateral 

 teeth, but young specimens do not show them. My sjiecimens of Castalina psam- 

 moica have the vertical ridges jioorly, or not at all, developed. 



This sculpture of vertical ridges upon the hinge-teeth in Castalia and the 

 allied genera has been much discussed in the literature. It is well known that 

 Neumayr believed that these ridges are homologous to similar structures seen 

 in the Trigoniidae, and that Castalia should be directlj^ connected with Trigonia, 

 and this idea has been taken up bj' other authors. However, Von Ihering (1893 

 p. 85) has pointed out that this is incorrect. From the above considerations we 

 now know, that there undoubtedly is a genetic series, which leads from Diplodon 

 through Castalina to Castalia, with the beginning at Diplodon, and the end at 

 Castalia; and, consequently, the vertical ridges on the hhige-teeth in Castalia are 

 not an ancestral character, inherited from Trigonia, but are a new acquisition, 

 marking an independent, comparatively high development of the Naiad-hinge.^" 



Hyria and Prisodon may, or may not, possess these vertical ridges on the hinge- 

 teeth. If present (^Prisodon alatus, PI. XL, figs. 2c, 2d), they are seen chiefly in 



'" This idea has also been suggested by Poini)cckj (Gegen Steiniiianu's (icologi.sche Grundlagen der 

 Alistanumingslohio, im ;5tcn Jahr.-Her. NiedersKchs. Geol. Ver., 1010, p. 12) in criticizing Steinmann's 

 attempt at a polyjjhylelic derivation of the Nniailrx from various Triijotiiidiv. Odhuer (191S, p. .')77, 

 footnote) also considers the sculpture of the teeth in ('iislalin as a secondarily developed character. 



