COMPOSITA 485 



extended posteriorly beyond the margin of the valve into the umbonal 

 cavity of the opposite valve ; the brachidium, including the jugum, with 

 its accessory lamellae similar to that of Atliyris. 



Remarks. This genus is usually more ovate or subquadrangular than 

 either Athyris or Cliothyridina, but its most essential differential char- 

 acter is found in the character of the surface markings, the lines of 

 growth being extended into neither continuous nor fimbriate lamellae. 

 These shells have been commonly designated under the generic name 

 Seminula since the publication of Hall and Clarke's 1 great work on the 

 genera of Paleozoic Brachiopoda, but Buckman has shown 2 that the geno- 

 type of McCoy's Seminula is a shell allied to Camaroplioria, and that the 

 name is consequently inapplicable to this group of shells, and he has 

 adopted the name Composite/, iBrown, which was based upon shells of this 

 type. 



The members of the genus compose an assemblage of forms between 

 which it is exceedingly difficult to draw sharp, specific lines. Those mem- 

 bers of the genus from the middle and upper Mississippian faunas of 

 the Mississippi valley basin are usually referred to one of two species, 

 C. trinuclea or C. sub quadrat a. The first of these was originally described 

 from the Salem limestone fauna of Indiana and Illinois, the second from 

 the Chester fauna of Illinois. An examination of a large number of 

 individuals from these two horizons, from both the original and other 

 localities, shows that series can be selected which pass with complete 

 gradation from one form to the other. Aside from these two commonly 

 recognized forms, an extended series of specimens shows several others 

 which if isolated would be recognized by any one as distinct specific 

 forms, but like the two already mentioned, they merge with complete 

 gradation into the general type. In addition to these forms the Penn- 

 sylvanian C. subtilita is united by regular gradations with these Missis- 

 sippian shells. Under these circumstances the student is forced to one of 

 two conclusions, he must either unite all of these forms into one exceed- 

 ingly variable species, or he must recognize several specific or varietal 

 groups which grade into each other. The last of these methods appeals 

 most strongly to the writer, and preference is given to the recognition of 

 the groups as of specific rank, there being little utility in giving a par- 

 ticular shell three names, when two will serve as well. Among 

 a group of any considerable number of individuals of these shells, from 

 a single horizon and locality, there is always a central form, a sort of com- 

 posite of all those individuals, represented by the average of the sum 

 total of characters, which is the true ideal of the species. It matters 

 not that a few of the individuals agree essentially in all their characters 



1 Pal. N. Y., vol. 8, pt. 2, p. 93. (1894.) 



2 Ann. and Mag. Nat. Hist, ser. 7, vol. 18, p. 325. (1906.) 



