AND PORTLAND OOLITE OF WILTS, ETC. 559 



include the perforate Thamnastrcece in the Poritidae or the Fungidae, 

 we at any rate can scarcely deny that the difference between per- 

 forate and imperforate septa is a character of sufficient importance 

 to be regarded as generic. 



But there are other peculiarities in the present species which 

 may be mentioned. The first is, that the imperforate septa are con- 

 nected by true synapticulae, and not by " oblique cross floors," or 

 tabular synapticulae ; and another is, the tendency to pass from a 

 turbinate to a digitate form, each of the finger-like processes having 

 its surrounding wall and terminal calicular surface. Specimens 

 having the divided form exhibit a distant resemblance to the genus 

 EpistrejotopliyUum of ILilaschewitch *, which genus is, however, 

 characterized by the presence of a single terminal calice, and by the 

 existence of both dissepiments and synapticulae. 



For the present I refrain from further remarks, excepting to 

 repeat my belief in the necessity of separating the imperforate Tham- 

 nastrcece from those which have their septa pierced, and the pores of 

 which have a distinct arrangement in the different species or groups 

 of species. 



Thamnastr^a coxclnna, Milne-Edwards and Haime, Brit. Foss. 

 Cor. pt. ii. p. 100. Plate XXII. figs. 1-4. 



Astrcea concinna, Goldf. Petref. Germ, i. p. 64. 



Yery massive specimens of this common species maybe met with 

 at many places. These often exhibit forms so remarkable as to 

 suggest specific or almost generic distinction, and they deserve 

 special notice. After the corallum has attained to a considerable 

 lateral growth, the upper surface throws up numerous conical pro- 

 cesses, which have their corallites directed outwards, just as they 

 are in ordinary dendroid Thamnastrcece. The corallites by their 

 outward and upward growth increase the diameter as well as the 

 length of the processes, until these, by pressure against each other 

 at their bases, become more or less angular or hexagonal, and are 

 separated only by narrow clefts or sinuses, which by the con- 

 tinued growth of the corallites become ultimately closed inferiorly. 

 When the upper surface of such specimens is afterwards worn down 

 to a convex, or more or less flattened form, what was the centre of 

 the conical processes still retains a certain degree of convexity, 

 owing to the position of the central corallites, which being vertical 

 are more able to resist the wearing process. A shallow depression 

 surrounds this convexity, while the line of union of the outwardly 

 directed corallites proceeding from the several processes, which 

 was previously a narrow cleft, now becomes a prominent ridge. 

 Such specimens, when cut through vertically, present the appear- 

 ance shown in figure 1 of the Plate accompanying this paper. In 

 these massive and worn specimens the well-developed and styli- 

 form columella becomes very conspicuous (fig. 4), and it is very 



* Palaeontographica, vol. xxi. p. 210. 



