160 PAPERS, ETC. 



mentioned, biit it hacl then assumcd the character and 

 development peculiar to the loop which characterizes the 

 genus to which it is referred. In an early stage of my 

 examlnatlon, modifications in the shape of the loop were 

 noticcd, and observations extending to several hundred 

 specimens rcsulted in showing the curious changes efFected 

 by age in the form of the loop, which may be seen by 

 referring to plate I., figs. 2, 3, 4, and 5. 



The first stage of development I liave been able to 

 observe is sketched in fig. 2. Therein it may be perceived 

 the two first lamellte are united to the hinge-plate, and to 

 a free rudimentary meslal plate, which is, in fact, the first 

 origin of what at a later pcriod becomes a mesial plate. 

 In this State it is free, and does not touch the bottom of 

 the valve, although when viewed in profile spines may be 

 seen passing downwards, which afterwards join the meslal 

 septum. 



The loop has not yet been formed, but a plate projects 

 between the lamellaj, and appears as if longitudinalJy split 

 to a certain depth in the centre. 



The second stage is exhibited in fig. 3. In this \ve find 

 the two lamella3 with the rudimentary plate as in fig. 2, 

 and, besides, the origin of the reflected portion of the loop, 

 presenting in this first stage of its development but a very 

 small and rudimentary aspect. 



The third stage may be observed in fig. 4, where the 

 different parts ai'e still more developed, but the mesial plate 

 has not yet reached the bottom of the valve. 



By gradual changes we are thus conducted to the fourth 

 stage, fig. 5, where the loop has attained its complete 

 development. The central plate, which was freely sus- 

 pended in the shell before, has now reached and become 

 soldered to the bottom of the valve ; the first pair of 



