124 PEOF. p. M. DUFCAK ON THE STRUCTFKE AlfD 



most certainly is not, it is attempted to alter the classificatory posi- 

 tion of Aocosmilia Wrightii, Ed. and H., and Montlivaltia Holli, nobis, 

 of the Inferior Oolite (E. F. Tomes, op. cit. p. 41 6). The two forms are 

 said to belong to the genus Donacosmilia, E. de From. That genus 

 may be thns defined from the ' Introd. a I'Etude,' p. 146 : — 



" Corallum fasciculate, reaching to a considerable height, wall 

 with a complete epitheca, which forms circular bourrelets ; 

 calicos circular. Septa broad, slightly exsert, and uniting in the 

 centre of the calico. jN^o columella. Dissepiments numerous and 

 well developed." In a description of one of the species, it is noticed 

 that the epitheca may wear and expose the septa to view close to the 

 wall, and that, when several bourrelets are thus formed one over the 

 other, it looks as if a set of corals had been placed one within the 

 calice of the other. 



Axosmilia Wrightii was thus named with considerable hesitation 

 by Milne-Edwards and Jules Haime, and they stated this and the 

 reason for their doubts in their great monograph. They never saw 

 the inside of the specimen, nor the outside of the calice, for it was 

 in matrix. Their figures show a short simple coral, marked not 

 overmuch by growth-rings, and there is nothing to lead to the 

 belief that the form was the beginning of a compound and fascicu- 

 late one. 



It would have been better if Mr. Tomes had stated the opinion 

 of the describers of Axosmilia in full. No one will admit that they 

 have been proved to be wrong until the absence of a styloid 

 columella is shown in Axosmilia Wrightii, and it is satisfactorily 

 demonstrated that the simple form was going to be a compound and 

 fasciculate one. A student of the recent coral-fauna will have no 

 difficulty in recognizing growth-rings in the excellent figures of 

 Montlivaltia Holli by fie Wilde, in the ' Monograph of the British 

 Fossil Corals,' 2nd ser. pt. iii. Pal. Soc. 1872, pi. i. figs. 5-8. The 

 coral resembles Donacosm,ilia in two points, but they are of no great 

 physiological importance. They are the resemblance of the outside of 

 the corallum to the base of a species of Donacosmilia, and the non- 

 existence of a columella. But the pedunculate nature of M. Holli, 

 and its shape, are common to some recognized Montlivaltice ; and 

 the nature of the axial part of the corallum, where a columella could 

 be formed, differs in JDonascosmilia and in M. Holli. In the one the 

 septa meet in the centre ; and in the other there is a considerable space 

 environed by the ends of the septa ; and thus there is no union. But 

 the distinctions between the form and any Donacosmilian are definite 

 and generic. There is decided calicular gemmation ; and this has 

 been mistaken for rejuvenescence, or rather for the results of that 

 process. Donacosmilia is a rare genus in the British fauna ; and 

 even on the Continent its species are found in a miserable condition. 

 Indeed, it was so doubtful whether there was not a columella in the 

 axis, that I placed a mark of doubt against it in the " Eevision "*. 



The nearest approach to anything resembhng a Donacosmilian 

 which has come under my notice, is a species of LepiclopTiyllia, 

 * " Eevision of the Grenera of Madreporaria," p. 54. 



