J. IMBRIE 149 



will serve as a convenient focus for a theoretical discussion of the 

 taxonomy of successional species. This review is based largely on 

 Kermack's (1954) critical restudy of Rowe's materials. 



From many horizons in the chalky limestones of the English 

 Cretaceous large numbers of well-preserved Micraster have been 

 collected. Most of the limestones are soft enough so that simple 

 preparation techniques make it possible to free the tests from the 

 enclosing matrix. Owing to the complexity of the echinoid test, a 

 large number of unit taxonomic characters can be distinguished 

 and interpreted functionally by analogy with living spatangoids. 

 From many points of view, then, these fossils make ideal materi- 

 als for evolutionary and taxonomic studies. 



Some of the outstanding features of the fossil record of Micras- 

 ter, as interpreted by Kermack, are shown in Fig. 15. The oldest 

 known members of the genus ( M. leskei ) possess small tests with 

 the widest portion of the test well ahead of midlength. The 

 mouth, lacking a distinct labium, is plainly visible in ventral as- 

 pect. The anterior ambital notch is shallow and the subanal fasci- 

 ole small. Closely spaced collections of the main line of evolution 

 from this form record a progressive shift in the average values of 

 the skeletal characters just mentioned, and in other characters as 

 well, giving rise to M. coranguinum. In this large species the 

 widest portion of the test is approximately at midlength, the 

 mouth is farther forward, and a strong labrum obscures the 

 mouth opening in ventral aspect. The anterior ambital notch is 

 deep and the subanal fasciole large. It should be emphasized that 

 distinctions between successive population samples can be made 

 only by comparison of the average values of overlapping fre- 

 quency distributions. 



Two side branches of the main M. leskei-coranguinum line can 

 be distinguished. The first, M. corbovis, is a large form with a 

 number of characteristic features not discussed here. The second 

 branch, M. senonensis, differs from its contemporary M. corangui- 

 num principally in its complete or nearly complete lack of a sub- 

 anal fasciole. Frequency distributions of this character in a single 

 population studied by Kermack show two nonoverlapping clus- 



