52 



TEOF. P. MAETIN DTJFCAN'S EEYTSION OF THE 



later Tertiary faunas. I do not consider Upismilia otherwise 

 than a subgenus. 



Subgenus Epismilia, (genus) ^. de Fromentel, Introd. d Vetude 

 des Fohjp.foss. p. 104 (1861). 

 Syn. Psammosmilia, E. de From. 



Coral] um free or adherent, trochoid or subcylindrical. Septa 

 large and entire. No columella ; columellary space elongate. An 

 epitheca, membraniform and complete, forming transverse ridges. 

 Endotheca abundant. 



Bistrihution. — Fossil. Jurassic: Europe. Cretaceous: India 

 and Europe. Eocene : Europe. 



It is to be noticed that in the typical species, Fpismilia Haimei, 

 E. de From., the septa are exceedingly numerous ; that the septa 

 of the 1st, 2od, and 3rd cycles have the upper free edge sharp 

 and entire, their inner edge being slightly swollen and undulating ; 

 and that the septa of the other orders have regular dentations 

 on their inner edge. The species have large individuals as a rule. 

 The only distinction bet^^een this genus and Trochosmilia, Ed. & 

 H., 1848, is that this last has the wall covered with costse, and a 

 rudimentary epitheca in some instances. This last structure is 

 not mentioDcd in the diagnosis of Milne-Edwards and Jules 

 Haime, but it is as evident as it is in some Caryophyllise. 



As it is of great importance that the genera should be well 

 defined, it is impossible to retain both of these ; and hence I place 

 Troc7iosmilia,^d. & H., 1848, as the genus which contains, amongst 

 others, certain species with a well-developed epitheca, belonging 

 to a subgenus, Fpismilia. 



Fsammosmilia, E. de From. Pal. Fran9., Terr. cret. 1863, differs 

 from Fpismilia in having a broad base, which is not generic. It 

 must be absorbed. 



The genus Coelosmilia, Milne-Edwards and Jules Haime, 

 has a considerable number of species with a well-defined facies, 

 but its alliance with Trochosmilia is exceedingly close. Formerly 

 I had absorbed Coelosmilia in Trochosmilia, but now I propose to 

 consider it as a subgenus. The main distinction between the 

 two types only relates to the amount of endotheca. 



Subgenus C(elosmilia, (genus) Milne-Fdwards Sc Jules Saime^ 

 Hist. Nat. des Corall. vol. ii. p. 175 (1857). 



The coralium is simple, fixed or pedicellate, subturbinate, 



