56 THE FOEAMINIFEEA 



possible to take a large series of a common form from 

 one locality and to arrange it gradationally. By this 

 means we can sometimes determine that the two 

 extremities of the series have as high a specific value 

 as many so-called species which occur but rarely. If 

 this can be demonstrated with a series of specimens 

 from a single locality, it is easy to conceive what an 

 enormous variation obtains with regard to what we 

 imagine to be a specific type over the whole of the 

 faunal area. Although this is true of many of the 

 Foraminifera, there are numerous type forms which 

 are well worthy of a specific name, for they present 

 marked characters of their own, and have shown 

 persistent characters throughout many successive 

 geological ages. 



Notwithstanding all that has been written against 

 the principle of using the character of shell-texture 

 as a prime factor in the classification of the Fora- 

 minifera, this method has very much to recommend 

 it ; and it also, besides its convenience, possesses 

 some merits for being a natural arrangement, so far 

 as it brings together certain tribes and types which 

 are clearly related to one another, even if they have 

 not emanated from the same prototype. 



The classification of H. B. Brady (1884) is based 

 on the earlier systems of Eeuss (1861), Carpenter, 

 Parker, and Jones (1862), and of Eupert Jones (1876). 

 The chief differences in Brady's classification are the 

 amalgamation of the families and the institution of 

 sub-families ; the creation of a new family, the 

 AsTROEHiziD.E, wMch received large accessions by the 



