AND GLYPHASTKiEA, DUNCAN (1887). 225 



Nothing is stated of the internal characters of the corallites of 

 this species, which, so far as can be determined from the figures, 

 is allied rather to the Lias forms wrongly placed in the genus 

 than to the type forms of Sej^tastrcea. The form is from the 

 Cenomanian of Le Mans, France. 



It would have inordinately increased the limits of this paper, 

 and would have involved an extended microscopic investigation, to 

 have determined positively the genera to which the Lias and other 

 corals belong, which I have shown cannot remain in SejptastrcBa. 

 I have further refrained from this task since I have learned, whilst 

 engaged on the present paper, that Mr. R. F. Tomes, F.G.S., 

 has undertaken an investigation of these Liassic species, of several 

 of which he is fortunate enough to possess the type specimens. 



Summary. 



The following are the main facts and conclusions arrived at in 

 the foregoing paper : — 



1. The genus Septastrcea, d'Orbigny (1849), is based on the 

 characters of a coral from the Miocene strata of Virginia, which 

 the author named S. subramosa, but did not describe. The same 

 form was subsequently named S. Forhesi, by Edwards and Haime ; 

 d'Orbigny claimed this name as the synonym of S. subramosa ; but 

 owing to this latter being merely nominal the claim cannot be 

 recognized, and the type species of Sejptastrcea must bear the name 

 8. Forbesi, Edw. & H. 



2. The original specimen of S.Forbesi, E. & H. (= /S'. subramosa, 

 d'Orbigny), is in the British Nat. Hist. Museum, and from it the 

 characters of the genus can be satisfactorily ascertained. 



3. In 1861 de Eromentel, and in 1867 Prof. Duncan placed in 

 the genus Septastrcea several species of corals from the Lias 

 Formation of England and France ; but these are not generically 

 related to the original types of the genus from the Miocene Tertiary 

 of America. 



4. Prof. Duncan, in 1887, takes out of Septastrcea the type 

 species S. Forbesi, and makes it the type of a new genus, Olyplias- 

 trcea^ leaving in the former genus those Liassic species placed therein 

 by himself and Fromentel at a later date, and which have no generic 

 relation to d'Orbigny's and Edwards and Haime's original types of 

 the genus. As the same species cannot be the type of two distinct 

 genera, the name Glyphastrcea becomes obsolete. 



5. In the genus ISeptastrcea, as based on S. Forbesi, amongst 

 other features enumerated, the corallites have separate waUs, the 

 theca is formed by an extension of the septal laminae, the lower 

 portion of the coraUite-walls and the septa are thin, but the upper 

 portion of the corallites is infilled with stereoplasm, so that the 

 calicos are shallow. There is no true columella, only a pseudo- 

 columella formed by the union and partial involution of the inner 

 septal margins. The increase is exclusively by marginal gemmation ; 

 fission does not occur. Linear perforations between the septa for 



