BRACHIOPODA. 2G7 



In the Devonian of North America we find species externally similar to 

 Terebratula Guerangeri, and shall presently recur to tliem in considering the 

 precise variation of structure which such forms exemplify. 



Further, there is evidence of the existence, in the Upper Helderberg 

 fauna, of a small, smooth biconvex terebratuliform shell, with sloping 

 cardinal margins which bears the same form of brachidium as Centroiiella 

 glans-fagea* 



This type of brachial apparatus, however different may be the exterior of 

 the shells bearing it, whether ribbed, striated or smooth, biconvex or plano- 

 convex, has a definite time-value, and our present knowledge indicates that, 

 beginning its existence in the Lower Helderberg, it is continued into the Oris- 

 kany sandstone, Upper Helderberg and Hamilton groups without material 

 modification. 



From this time onward the structure yielded to modifying influences. In 

 the Centronella Julia of A. Winchell, a smooth, biconvex species from the Mar- 

 shall group of Michigan, at the base of the Lower Carboniferous series, the 

 anterior triangular plate of the brachidium bears a high vertical lamella extend- 

 ing in both directions beyond the limits of tluit plate into the cavity of the 

 pedicle-valve. 



Such forms must be kept apart from the earlier and typical Centronellas, as 

 indicative of a progressed stage of development. The strongly plicated shells 

 from the Productus limestone (Carboniferous) of the Salt-Range of India, which 

 Waagen has described under the generic name Notothvris (though the struct- 

 ure of their brachidia is not fully known), will undoubtedly prove to possess 

 some modification of the Crntronella type. Such a modification is evident in 

 Derby's Notothyris ? Smithi {loc. at.) from the Devonian, a similarly plicated 

 species which may prove to be an ancestral form of the Carboniferous 

 Notothyris. 



Returning again to the genus Centronella and restricting our conception 

 to forms similar to that regarded by its author as typical, the following diag- 

 nosis may be deduced : 



* Farther I'eference will be made to this shell. 



