THE NAUTILUS. 119 



similarity of the anatomy and shell, that a tasidium is attributed to 

 all Mutelidde. Nevertheless, it is quite probable, that Spatha has 

 also a lasidium, for the South American genus Glabaris is still more 

 like Spatha in its structure, than is Hyria and its allied forms. I 

 have investigated a number of specimens of Glabaris, Fossula, and 

 Monocondyltza, and they all had the essential structure of Spatha : a 

 firm bridge separating anal and branchial opening ; the same attach- 

 ment of the anterior end of the inner gill ; the marsupium practically 

 occupying the whole length of the inner gill ; and the gills with 

 well-developed water-tubes and heavy, continuous septa. The differ- 

 ences from Spatha are found in the connection of the inner lamina 

 of the inner gills with the abdominal sac, which I consider unim- 

 portant ; in the shape of the palpi, which are rather variable, but 

 always more or less rounded and broad, not falcate ; and in the anal 

 opening, which in Glabaris, Fossula, and Monocondylcea, is always 

 entirely open and closed nowhere. The most interesting fact in the 

 three latter genera is the similarity of the structure of the water- 

 tubes and septa of the gills to that of Spatha. If we may say that 

 Hyria, Tetraplodon and Diplodon ought to stand in the same family 

 with Spatha, we must also admit Glabaris, Fossula and Monocon- 

 dylsea, and the latter are even more closely allied to Spatha than 

 the former. Thus it would be suggested that we have here one 

 family, possibly to be called Mutelidse, which is divided chiefly by 

 the character of the gill structure and rnursupium, but also by shell 

 characters (hinge), 1 into two subfamilies, the one (tfyriinaef) includ- 

 ing at least three South American genera, the other including three 

 South American and one African genus. For the latter possibly 

 the name Mutelince might be used, but this depends on our knowledge 

 of the soft parts of the type-genus Mutela. 



The conclusions to be drawn from the present investigations are: 



1. Of the Asiatic so-called Hyninte at least Parreysia and Lamel- 

 lidens are to be removed, and are to be associated with the North 

 American genera Qaadnda, Rotundaria, Pleurobema and Unio in the 

 subfamily Unionince of the family Unionida. 



2. The African Mutelid genus Spatha and the South American 



1 Also the larva (lasidium) might furnish a differential character. I have 

 seen several gravid specimens ot Glabaris, but they had only eggs, and no fully 

 developed larva. 



