226 THE AMERICAN MONTHLY [Aug., 



all of which, remember, are to be found in the Barbadoes de- 

 posit. But these 191 genera are altogether too much for one ar- 

 ticle. However, they fall conveniently into two groups in one 

 of which there are 71 genera. Some of these were classified in 

 my first paper, but more were not. I said the Peripyla?a (or 

 Spumellaria) contained 20 genera and I gave ihe names. Two 

 of those genera I find now are not represented in Barbadoes, 

 namely, Stauradarn and Echinactura. So that of the 71 genera 

 only 18 have been touched upon, in other words 53 will be en- 

 tirely new. Under these circumstances I think it will be wise 

 to begin all over again. 



Let us take then the 71 genera of the Spumellaria, or Peri- 

 PYLiEA., in which the skeleton is round or disk-like and has no 

 large mouth or opening and the central capsule is spherical and 

 has pores distributed all over it. 



Now these 71 genera are divided into 3 main groujis, thus : 



I. Sphseroidea — Spherical Shell. 



II. Prunoidea — Ellipsoidal Shell. 

 IIL — Discoidea Discoidal Shell. 



Taking the first in order, we have 21 genera arranged accord- 

 ing to the presence or absence of radial si)ines and the num- 

 ber of spheres. 



L— SPHERICAL FORMS. 



1.— Family : LIOSPH.ERIDA. No radial spines. 



A, — ONE SPHERE. 



Tulnili. Ethmosphsera. 



None. Cenosjjhxra. 



B. — TWO SPHERES. 



One within capsule. Carposphasra. 



C. — THREE SPHERES. ' 



One shell within capsule. Rhodosphaera. 



Two shells within capsule. Thecosphcera. 



D. — FOUR SPHERES. 



Two inner medullary, two outer cortical. Oromyoaphcey-a. 



2.— Family: STYLOSPHiERIDA. Two radial spines op- 

 posite in one axis. 



A. — ONE SPHERE. 



Spines equal. Xiphosphcera. 



Spines different. Xiphostylus. 



