CORALS OF THE CIXCIJ^TNATI GROUP. 185 



It may be a question, however, as suggested by Milne Edwards and 

 Haime, if the form described originally and carefully figured by Goldfuss 

 under the name of Columnaria alveolata (Petref. Germ., pi. 24, fig. 7), be 

 not truly an example of Favistella stellata, Hall — the specimens seen by 

 Goldfuss being said to come from Seneca Lake, and being, therefore, 

 probably derived from a traveled bowlder. The form now universally 

 recognized as Columnaria alveolata, and so characteristic of the Trenton 

 limestone, unquestionably has the septa rudimentary and marginal, in- 

 stead of being well developed and reaching to the center of the corallites, 

 as represented in the figures given by Goldfuss. In any case, however. 

 Hall's genus Favistella may be advantageously retained for corals which 

 resemble Columnaria, as now accepted, in most essential respects, but 

 differ in the completeness of the septal system and the more thorough 

 amalgamation of the walls of adjacent corallites. 



The only species of Favistella which ajDpears to occur in the Silurian 

 rocks of Ohio is F. stellata, Hall, which is likewise an abundant species 

 in strata of a corresponding age (Hudson River group) in Canada. 



Favistella stellata, Hall. 



Favistella stellata, Hall; Pal. N. Y., Vol. I., pi. 75, figs, lo-lc. 



Corallum sub-hemispheric, or pyriform, massive ; corallites prismatic, 

 usually hexagonal or pentagonal, from rather more than one line to two 

 lines in diameter, completely amalgamated with one another by their 

 walls, except occasionally towards the surface of the mass. Increase by 

 fission of the old tubes. Septa unequally developed, alternately large 

 and small ; large septa usually fifteen, sometimes fourteen, in number,, 

 reaching to the center of the visceral chamber, or nearly so; small septa 

 marginal and rudimentary. The total number of septa is thus from 

 twenty-eight to thirty. Tabulae well developed and complete, about 

 three in the space of one line. No columella. 



According to Hall, there are usually twelve septa, more or less, in this 

 species; but in this enumeration he has evidently counted only the 

 large septa and disregarded the small ones. The number of large septa, 

 however, is usually fifteen, and the rudimentary secondary septa are 

 placed alternately with the former, and are, therefore, the same in number. 

 In this respect, all the specimens I have examined appear to agree. In the 

 Hudson River group of Canada occurs a species of Favistella, which agrees 

 in most respects with F. stellata, Hall, but which exhibits the peculiarity 

 that in many examples the corallites are not amalgamated with one 



