or SOUTHERN INDIA. 
23 
a mucli more restricted sense to the Astreidm ; for, according to the above noticed 
characteristic, it will also include his Sympjiylliem, Fayieye, 
and two or three others. It is possible, and I should say very desirable, that a few 
separate groups or families should be distinguished ; hut, for reasons already referred 
to, I hardly think that this can he done with advantage in the form introduced by 
Fromentcl. It is often, in imperfect specimens, very difficult to decide in wdiat 
manner the multiplication of the corallites takes place, and until all these and other 
points in structure have been satisfactorily settled, it will be, I think, for our pur- 
pose more advantageous to retain the extent of the AstrrieA': in the older sense. In 
conformity with this, the family is divided into two groups according to whether the 
multiplication of the corallites takes place by fissiparity, or by suh-marginal gemma- 
tioir ; the former may ho called symruyllixa!:, the other astreixye (excluding the 
Tiiamxastreieye). Between these two sub-divisions come the fafiixas. 
The SYMEiiYLLix^ ai'c represented in South India by the following genera ; 
Thecosmilia Avith one, Jloloccenia Avith two, Astroocenia with four, ILjcelopliyllia 
with two, and Stellovia with one species ; the astreix^r have only four genera, 
IleUastrea Avith three, Flaoastrea (n. gen.) with one, Isastrea with five, and 
Lalhnceandra also Avith five species. Thus Ave have in all tAventy-four species of 
Astreiem, hut tAvo or three remain somewhat imperfectly knoAvn. 
X. Genus.' — THECOSMILIA, Milne-Fdwards and JIaime, 1818. 
Hist. Nat. lies Corall., vol. ii, p. -Sol. 
The corallum is composite, emspitose, sometimes forming large aggregate 
masses connected together by a well developed epitheca, hut the single iudividuums 
generally become free hear the calyx, and the epitheca reaches very nearly or fully 
to its edge ; the calyx is slightly impressed in the centre Avith somewhat irregular 
edges ; the septa, arranged in six systems, are laterally distinctly granular ; in 
typical specimens no columella is present, hut a few aberrant forms have traces of 
it ; the endotheca is abundant. 
All the species, as far as knoAvn, are, I believe, fossil, and the majority from 
mesozoic deposits, particularly cretaceous and Jurassic. 
Tukcosaiilia gemixata, StoUezka. PL IV, Eigs. 1 — 3. 
Thecosm. corallum hreviler ccespilosum, hi- vel pohj-geminalum, calgcibus suh- 
rotundatis, allingenlibus, theca crassa comuni circumdatis, in junioribus fere planis, 
in adtdtis mod ice convexis, medio paulo concaviusoulis, epUheca usque ad marginein 
calycis extensa ; septis crassis in quatuor aut quinque cgclis disqwsilis. 
This is a very remarkable species, in some respects intermediate between Fu- 
phyllia and Thecosmilia ; the presence of a strong epitheca appears, hoAAmver, to 
shoAV a decidedly greater relation to the latter genus, though in other respects it also 
( lo5 ) 
