102 THE NAUTILUS. 



UnionidcB fully bear out Sferki's division into two groups : summer 

 and winter breeders. The breeding season of the summer breeders 

 is short (maximum hardly four months), while in the winter breed- 

 ers this season is prolonged, extending from late summer, through 

 the winter into spring. Yet it must be borne in mind that probably 

 in the single individual the breeding season does not fully occupy the 

 whole length of the term, since it has been repeatedly observed that 

 the embryos and ovisacs are discharged at various times, even in the 

 beginning of the winter. 



In some species belonging to the group of winter breeders the 

 period of gravidity may be extremely long, so that the end of one 

 breeding season (in May, June, July) may overlap with the begin- 

 ning of the next (June, July, August), and such species may appear 

 to breed "all the year round." This has been hinted at already by 

 Sterki, and Conner gives the following instances: Lampsilis radiata 

 and Lampsilis nasuta. My own observations make this condition 

 probable in Lampsilis ventricosa and Lampsilis luteola. Yet in 

 others an " interim " is very distinct in the early summer. This is 

 the normal condition, according to Sterki, and has been found to be 

 true for Anodonta cataracta by Conner, and by my observations it is 

 made more or less probable for Lampsilis ligamentina, Ptychobranchus 

 phaseolus, Strophitus undulatus, Anodonta grandis, Anodovtoides fer- 

 ussacianvs, Symphynota compressa, Symphynota viridis, Symphynota 

 costatu, Alasmidonta marginata. 



These peculiar conditions may be explained by the following 

 assumption : Quadrula, with the four gills serving as marsupium, is, 

 in my opinion, the most primitive type of our Unionid(e. Next to it 

 stand Pleurohema and Unio^ with only the outer gills serving as mar- 

 supium, but with the shell more or less resembling that of Quadrula. 

 These forms represent also the most primitive type of the breeding 

 season, which is short, and falls into the warm season. These forms 

 existed already at a time when a uniform warm climate prevailed. 

 At that time, possibly, the breeding season was not so restricted, but 

 at the present time it has become so, since only during a short period 

 of the year these old, primitive conditions prevail (in summer). 

 Forms like Unio and Quadrula actually go back to mesozoic times. 



All other genera are more advanced. The group Alasmidonta, 

 Symphynota, Anodontoides, Anodonta, Strophitus (which is, accord- 

 ing to Sterki, characterized by a peculiar glochidium) resembles in 



