52 THE NAUTILUS. 



corresponds to my subfamily Anodontinae (Ortmannn, Naut. 

 23, 1920, p. 117 and Ann. Carn. Mus. 8, 1912, pp. 224 and 

 278). However, it does not belong here at all. It is true, the 

 rudimentary condition of the hinge suggests its affinity with 

 the Anodontinae, and what Simpson knew about the anatomy 

 did not conflict with this. But the material at hand proves 

 conclusively that none of the characters of the Anodontinae are 

 present. It is, indeed, a member of the family Unionidae (as 

 defined by myself 1. c. ), for it has a complete diaphragm formed 

 only by the gills ; it has a supraanal opening ; the gills have 

 septa and water tubes running parallel to the gill filaments. 

 However, the facts, that all four gills are marsupial ; that the 

 charged gills are only moderately swollen, with sharp edges ; 

 that no system of secondary water canals is developed within 

 the gills ; and that the glochidia are not triangular and have no 

 hooks, place Gonidea with the subfamily Unioninae. 



Within this subfamily, the genus has quite an isolated posi- 

 tion, offering a curious mixture of primitive and advanced 

 characters. The most primitive features are, that all four gills 

 are marsupial in the female, and the interrupted character of the 

 septa. This latter character is quite unique, suggesting even 

 the ancient family of the Margaritanidae. The rather long 

 mantle-connection between anal and supraanal openings does 

 not agree with the more primitive types of the Unioninae (Fus- 

 conaia etc.), but rather with the more advanced ones {Unio, 

 Elliptio), while the rudimentary condition of the hinge again is 

 unique in the subfamily, exhibiting an advanced condition, 

 which is not known, except in the genus Lastena (see Naut. 28, 

 1915, p. 106). The simple beak sculpture (4 to 5 subconcentric 

 bars) appears as primitive. The glochidia are also of a primi- 

 tive shape, agreeing with the shape generally found in Unioninae. 



Hannibal (Science, 36, Dec. 20, 1912, p. 865) has suggested 

 an amended division of the Nayades into families and subfami- 

 lies. I am not prepared to accept this as proposed, but I 

 believe we shall be finally compelled, chiefly for the sake of 

 convention, to follow his fundamental idea, namely that my 

 subfamilies ( Unioninae, Anodontinae, Lampsilinac) should rank 

 as families. Then my Unioninae would become Unionidae (not 



