CORALS FROM THE CORAL RAG. 81 



All the specimens which we have met with were worn, and had lost their wall as well 

 as their basis ; we are, therefore, unable to decide whether this species was free or adherent, 

 and had or had not a complete cpitheca. There remains, therefore, some uncertainty relative 

 to the zoological affinities of this fossil, but we have referred it to the genus Montlivaltia 

 rather than to the genus Trochosmilia, on account of its great resemblance to some other 

 corals which undoubtedly belong to the genus Montlivaltia, and also because we have as 

 yet not met with any species of Trochosmilia in deposits formed before the cretaceous 

 period. 



The ^cww^ Montlivaltia, established by Lamouroux, contains a great number of species; 

 more than thirty have been described in our Monograph of the family of Astreidas, but 

 many of them are as yet but imperfectly known. In a note published a short time ago/ 

 M. D'Orbigny has considered it advisable to form a separate generic division for the species 

 which are of a compressed form, and he has given the name of Perismilia to the group 

 thus characterised. But this innovation is not, in our opinion, judicious, for, independently 

 of there being instances of every intermediate degree between the species with a calice 

 perfectly circular (such as Montlivaltia brevissima), and those in which the great axis of 

 the calice is to the short axis as 2G0 : 100, we see no reason for establishing generical 

 divisions on a character which, although to a certain degree constant in some cases, is in 

 others variable in the different individuals belonging to the same species. We must add 

 that no important difference in other parts of the corallum corresponds with the modi- 

 fications in the form of the calice. 



Since the publication of the first part of the Monograph, we have been enabled to 

 examine a great number of w^ell preserved specimens of Montlivaltia, and have been thus 

 led to rectify an error which the study of imperfect specimens had led us into ; we have 

 ascertained, in many species, that the edge of the sejjta is not entire, as we formerly sup- 

 posed, but is crenulate or regularly denticulated. There is, therefore, no longer any reason 

 for separating from the genus Montlivaltia the group which we established some years ago 

 under the name of Thecophi/lUa^ and the genus Montlivaltia, thus extended, must no 

 longer be placed in the section Eusmilinse (p. xxiii), but be referred to the tribe of the 

 Astreinas (p. xxxi). 



Montlivaltia dispar differs from M. deltoides,^ M. riidis,^ M. cornucopia^ M. hilohata^ 



1 Note sur des Polypiers Fossiles, 1849. 



2 Compt. Rend, de I'Acad. des Sc, t. xxvii, p. 491, 1848. 

 ^ Anualcs des Sc. Nat., s. 3, vol. x, pi. 6, fig. 3. 



'' C]/atItopJiyUum rude, Sowerby, Trans, of the Geol. Soc, s. 2, vol. iii, pi. xxxvii, fig. 2. 



5 Milne Edwards and J. Ilaime, loc. cit., p. 298. M. D'Orbigny places this species in bis genus 

 ElUj)sosmilia, (Note sur les Polypiers Fossiles, p. 5,) which is composed of our compressed Trochosmilia ; 

 all the specimens known are in a very bad state of preservation, but v.e are inclined to think that this fossil 

 had a complete cpitheca, as is the case with Mo«<//i'«/^('a; if not, it must be referred to owy ^enus, Trochosmilia, 

 for the subdivision of which, proposed by M. D'Orbigny, does not appear to rest on sufficient grounds. 



^ Turbinolia hilohuta, Michclin, Icon., pi. Jxii, fig. 1 ; (not pi. Ixi, fig. ".) 



11 



