468 MEMOIRS OF THE CARNEGIE MUSEUM. 



the anterior half of the gill, and less than one-fourth of the posterior end non- 

 marsupial. 



Thus it is clear, that, while the Castalia-Hyria-group of genera in the shape of 

 the shell and the hinge-teeth, and the C a stalia-gr on\) in the conformation of the 

 branchial openings represent a higher specialisation, in the rest of their anatomy, 

 and chieflj' in their marsupial characters, they all remain very close to the primitive 

 Diplodon-type. 



The Glochidia. 



(See Text-figure 4.) 



In conclusion I may say that I have found differences in the glochidia within 

 the genus Diplodon. That certain South American species of " Unio" have glochidia, 

 was first announced by Lea in the case of U. peculiaris and U.firmus (Lea, Obs., XII, 

 1869, PI. 34, figs. 80, 82; first published in 1868). He describes and figures them as 

 subtriangular in outline, obliciue, or upright, the ventral margin with a point, and in 

 the text he says that they are "furnished with hooks." However, he neither 

 describes nor figures these hooks. 



Von Ihering (1893, ]^. 47) states that hooks are missing in all of the South 

 American species examined by liim, but that the larvae are true glochidia. 



This is about all we have known hitherto about the larval form of the South 

 American "Unios" = Hyriimv. I have now found that the normal shape (out- 

 line) of the glochidia of Diplodon, and of the subfamily Hyriincc, is as described bj' 

 Lea, namely subtriangular, with a point on the lower margin (see fig. 4). In addi- 

 tion, I have found that some species actually possess a "hook;" but this hook is 

 entirely different from the one so well Icnown in the European genus Unio and in 

 the Anodontince of Eurasia and North America. The very fact that Lea mentions 

 the hook in a kind of perfunctory way, not calling attention to its peculiar features, 

 suggests that he never saw the real hook, and that he simply took the point of the 

 lower margin of the glochidium for it. 



This Hyriine hook (Fig. 4, h, c, d, e, f, h, k, I, m) differs entirely from the Ano- 

 dontine hook. The latter is triangular, attached by a broad base to the point of 

 the lower margin, and carries upon its upper surface a number of fine spinules. 

 The Hyriine hook is long and narrow, spiniform, with very narrow base, articulated 

 to the point of the lower margin, and without any spinules on the upper face, and 

 furthermore has a peculiar S-shaped curve. 



It is perfectly clear that this hook is different from that found in the genus 

 Unio and the Anodontince. Functionally it may serve the same purpose, that of 



