180 E. p. TOitES ON ilADEEPORAEIA 



inch in heiglit, which are usually attached to some hard body, and 

 over which they spread, in a manner which is the result either of basal 

 gemmation, or of increase by means of stolons. Of several speci- 

 mens now before me^ one from Parley Down is attached to a weathered 

 fragment of Thamnastrcea, which it partly conceals ; and another, 

 from the Lime-kiln quarry near Cirencester, is parasitic on Convex- 

 astrcea Waltoni, over which it spreads. 



The corallites are small at their attachment, rather crooked, 

 increase in size gradually, and attain to three fourths of their full 

 height before freely branching out. This they do irregularly, though 

 lateral gemmation apparently takes place about the same time in 

 the several coraUites. Once commenced, it proceeds with sufficient 

 rapidity to form a crowded mass of coraUites, the calicular extremi- 

 ties of which, however, have every degree of prominence, scarcely 

 any two being of exactly the same height. 



The walls of the corallites are thick, but do not extend to the 

 margin of the calicos, and there is no appearance of epitheca, but 

 well-marked costse which blend with the septa pass down the coral- 

 lites, and become less distinct as they extend downwards. 



The calices are circular and prominent, but there is a well-marked 

 fossula. The septa are exsert and are continuous outwardly with 

 the mural costae. The first two cycles are well developed and pass 

 into the columella ; the third cycle has very short septa, which are 

 scarcely a fourth of the length of those of the earlier cycles. The 

 columella is large, styliform, and with a pointed apex, but is not 

 very prominent. 



Height of the coralLum about 9 lines ; diameter of the corallites 

 about one line. 



A specimen from Earley Down indicates, by the manner in which 

 the corallites spread over the substance to which they are attached, 

 the possibility, almost the probability, of increase by means of stolons. 

 Some other examples, taken from the Lime -kiln quarry, Cirencester, 

 present precisely the same peculiarities. 



Styxosmilia excelsa, n. sp. (Plate Y. figs. 9-12.) 



The coraUum consists of a dense mass of slender corallites, and very 

 closely resembles those of Calamophyllia racUata, so closely, indeed, 

 that the two species have been confounded. On nearer examination 

 the corallites are seen to be a little less straight than in that species, 

 and rather more densely placed. There is no trace of the long 

 mural costae which characterize Calamopliyllia, the corallites appear- 

 ing to be quite smooth. The walls are thick, and the calices appear 

 to have been shallow, but none of them retain their margins. There 

 are about 12 septa, which are thin and not very regularly developed. 

 In some calices those of the third cycle join the older ones near to 

 the columella, but more frequently all of them run into the columella; 

 the latter is large and styliform. In some of the calices two opposite 

 septa are much thicker than the others, which is suggestive of fissi- 

 parity. 



The only specimen I have seen is of considerable size, and was 



