112 



entire, and it could not be said whether the third order was present in 

 all the systems or not. Costae, none visible ; but the surface is covered with 

 fine grooves which divide it into small polygonal spaces on which there 

 are faint granulations, but the whole structure is so much worn that 

 this appearance is exceedingly difficult to make out. Dimensions : — 

 Length of the branches from '25 to "37 ; diameter, from *3£ to *7 ; 

 distance between the calices on the same side 8 millimeters. 



Family Astr^eid^e— Group Trochosmiliaceje. 



Genus Conosmilia {Duncan), 1865. 



In the descriptions of new corals from Australia, published by Prof. 

 Duncan in the " Annals of Natural History for 1865," (p. 182), a new 

 genus was erected for simple pedicellate corals with a twisted laminated 

 columella and scanty endotheca. This was Conosmilia. I have given a 

 synopsis of the genus in the " Proceedings of the Royal Society of N. 

 S. Wales " for last year. I think the general character of these fossils 

 may become the type of a family, as there are other anomalous genera 

 to be noticed presently. In the collection there is one species of what 

 I believe to be this genus ; but it is a deformed specimen, and therefore 

 any attempt to determine the details would be hazardous. I shall merely 

 give it a name and some of its leading features, and leave the rest until 

 better specimens are found. 



Conosmilia contort a, n. sp., pi. 1, fig. 3. 



I 



Corallum, very much twisted and distorted, but normally a very 

 elongated one, cylindrical or slightly elliptical. Costae, broad, flat, or in 

 places acute, smooth or indistinctly vermiculate, corresponding with 

 septa. Epitheca, thin, smooth, somewhat shining, in concentric folds 

 occasionally. Septa, not granular, not exsert, but projecting in rounded 

 edges more or less according to the orders. Systems, six. Cycles, four. 

 Fossa, rather deep. Columella 1 Endotheca, scanty. Long. 20 lat. 

 of calice, 8. One specimen with doubtful fragments of two others. 



In the new species I am about to describe we have Conosmilia with- 

 out a columella, but with pali. In every other respect it resembles the 

 above-named genus of fossils, which Professor Duncan was the first to 

 describe. The absence of a columella or the presence of pali has always 

 been regarded as of generic importance in the present classification. I 

 therefore propose a new genus for these corals, which I shall name Cya- 

 thosmilia. There is considerable analogy between this genus and Cono- 



