AND GLrPHASTEiEA, DUNCAN (1887). 217 



existence. At present there is great confusion, and some of the lead- 

 ing authorities * differ widely even as regards the characters which 

 constitute a coral. 



From the foregoing description of the type species, Septastrcm 

 Forhesi, it is evidently necessary to revise the definition of the genus 

 given by D'Orbigny'and by Edwards and Haime, and I therefore 

 propose the following diagnosis : — 



MADREPOEAEIA APOEOSA. 



Family Aste^id^. 



Genus Septastrjea, d'Orbigny (1849), emend. Hinde. 



Syn. SeptastrcBa, d'Orbigny ; Edwards & Haime, in part ; Fro- 

 mentel, in part; Meek; Zittel ; non Duncan. Glyphastrcea, 

 Duncan. 



Generic Characters. — Corallum compound, massive, palmate or 

 subramose. Corallites prismatic at surface of corallum, subcylin- 

 drical in the interior, their walls in close contact with, but distinct 

 from each other. Septa in normal corallites varying from 6 to 24, 

 usually either 6 or 12 fully developed ; septa of the third cycle not 

 reaching the centre of the calice. Occasionally two opposite septa, 

 much more prominent than the others, meet in the centre and ex- 

 tend across the calice. The septal laminae extend laterally at the 

 periphery to form the theca. The inner margins of the septa unite 

 with each other to form closed loculi, and they are occasionally in- 

 tertwisted, so that a pseudo-columella is produced. Dissepiments 

 nearly horizontal. The septa and thecal walls are thin and delicate 

 in the lower part of the corallites ; but in the upper portion, when 

 full-grown, the interior of the corallites is filled with a deposit of 

 stereoplasm, so that the calices are very shallow, and in mature 

 forms there is a complete floor to the calice. Increase by gemma- 

 tion in the interspaces between the corallites. 



Type species: JSejotastrcea Forbesi, Edw. & H. ( = S. suhramosa^ 

 d'Orbigny, nominal). 



Loc. Fossil : Miocene Tertiary, North America. 



The Mesozoic forms placed in the genus by de Fromentel and 

 Duncan do not properly belong to it. 



As the mode of increase of this genus is clearly by gemmation, it 

 can no longer remain in the tribe of the Faviacege, in which it was 

 placed by Edwards and Haime, nor in the alliance Goniastraeoida of 

 Prof. Duncan. 



* As an instance of this may be cited the fact that whilst Mr. J. J. Quelch 

 merges the Madreporaria Rugosa as a single group with the Madreporai-ia 

 Aporosa ('Challenger' Eeport, vol. xxi. 1886, p. 57), Prof. P. M. Duncan states 

 that it might be regarded as doubtful whether the Rugosa are Corals at all ! 

 (Quart. Journ. Geol. Soc. vol. xl. 1884, p. 506). The same author, again, in 

 treating of the Fungidae, writes, "One might speculate upon the impossibility 

 of the occurrence of mesenteries, and wonder whether these forms are really 

 corals " (Linnean Soc. Journ. vol. xvii. 1883-4, p. 146). 



Q. J. G. S. No. 174. Q 



