446 PROCEEDINGS OP THE GEOLOGICAL SOCIETY. [May 6, 



part Astraeans. By far the greater number belong to the species 

 I have named Astrcea crassolamellata, from the great thickness 

 of the outer part of the principal septa. The six varieties and the 

 typical form of this species form a most interesting group; and 

 although no specimens have as yet been discovered except in Antigua, 

 still there are some remarkable specific alliances. I have preferred 

 terming the several forms described, varieties, for they all have the 

 closest structural resemblance to each other and to the typical form. 

 There is generally some variation in the appearance of the indi- 

 viduals of compound corals ; this is not simply the result of imma- 

 turity or of arrested development, but is to be referred to the causes 

 which determine individual differences in secondary and tertiary 

 characters. Some masses of the species Astrcea crassolamellata con- 

 sist of aggregations of corallites which have those individual peculi- 

 arities and deviations from the type which are to be seen in a few 

 corallites in the typical specimens ; and in the variety pulchellu, 

 which is very polymorphic, the corallites of a large corallum vary 

 sufficiently in different parts of it to constitute other varieties were 

 the specimen broken up and its original entirety not understood. 



The restriction of Astrcea crassolamellata to the Marl-formation 

 is interesting. The species does not appear to have any close "West 

 Indian affinities, either fossil or recent, and it must be referred to 

 that part of the Miocene Coral-fauna which, having a much earlier 

 origin, died out in Tertiary times. 



The fossil Corals of the "West Indies appear to have relation to 

 two distinct Coral-faunas, one of which existed during the Oolitic 

 period, and has also left its traces in the Lower Cretaceous strata, 

 and another whose first appearance is uncertain, but which attained 

 its greatest development during the Miocene period, and is now 

 represented in the Pacific Ocean and its associated seas. 



The European fossil Corals related by structural resemblances to 

 Astrcea crassolamellata are Astrcea Lifolensis, Middle Oolite, and 

 Astrcea Guettardi, Miocene. A specimen of the first from the Yosges, 

 in the British Museum, has great resemblance to those of the new 

 species; and the description of Astrcea nobilis, now merged into 

 Astrcea Guettardi by Milne-Edwards and J. Haime, suggests the 

 existence of several important points of structural affinity between 

 that species and Astrcea crassolamellata. 



The species Astrcea endoihecata, of which there are varieties in 

 Antigua, is a San-Domingan type; and Astrcea Antillarum, from 

 Montserrat, is closely allied to it. 



Astrcea cellulosa and its variety with curved septa are very 

 remarkable forms ; they are in many respects miniatures of the 

 great Astrcea endothecata, and are found in the older strata. Their 

 discovery in a formation where fossil "Wood, and Testacea not purely 

 marine, are noticed is interesting ; for the presence of fresh or 

 brackish water is very antagonistic to coral-growth, and especially 

 as the great Astrcea which they resemble was found in the Marl, 

 which contains no evidence of such conditions. 



Astrcea megalaxona increases by fissiparous growth, as well as by 



