A. E. Verrill — Bernnidian and West Indian Reef Corah. 1 17 



I can find no evidence of more than three species, and I am not cer- 

 tain that more than two of these can eventually be kept apart. 

 <^uelch, however, with a much smaller series, recorded (op. cit., pp. 

 10, 11) eight species from Bermuda, including the young forms that 

 he called LithophylUa. Probably all his Bermuda forms belong to 

 I. dipsacea and /. frag His. 



When young, all species of this genus and of Massa (including 

 :Sy}nphyllia), etc., have a simple, more or less cup-shaped coral, 

 attached by a rather broad base. These may become in some cases 

 2b to 40""" in diameter before they begin to form marginal infold- 

 ings, as a commencement of the process of fission. 



Such simple young forms have been put in a special genus 

 {Scolymia Haime, 1852, or LithophylUa Edw. and Haime, 1857). 

 The type of this genus was M. lacera Pallas. It appears to be the 

 young of Mussa carduus (Ellis and Sol., sp.).* 



Therefore LithophylUa is a synonym of 3Iussa, rather than of 

 Isophyllia, though several species described by Duch. and Mich, 

 unquestionably belong to Isophyllia, as indicated in the synonymy 

 above. All the Bermuda simple forms are young of Isophyllia, and 

 mostly of I. dipsacea and /. frayilis. PI. xix, tig. 5. 



The generic relations of these simple young forms can usually be 

 told by the character and spiiiulatiou of the costae. In Mussa the 

 costte are generally imperfect, with rows of strong, sharp spines, often 

 recurved. In Isophyllia the costte are generally raised and continu- 

 ous ribs, often lamelliform, and their spines are small and more 

 regular, usually more like serrations of the edge. In Mussa the 

 septa are also more strongly and more unevenh^ serrate or lacerate, 

 especially toward the outer end. 



At a later stage, but varying in size, even in the same species, the 

 edge of the cup begins to be undulated or lobed ; most commonly 

 there are six otitfoldings and six infoldings at first, corresponding to 

 the primary and secondary septa, but the nunibei" may vary from 

 three to seven, or even eight or more. When four lobes are formed 

 the coral is apt to be squarish. (See pi. xvii, tig. 4.) These primary 

 folds and lobes may continue to grow regularly for some time, till 

 several large marginal calicles, usually five or six, develop around 

 the central, stellate, primary calicle (pi. xvii, figs. 1-2). This is the 

 most normal and regular mode of growth for all the species of this 



* This large species should, therefore, be called Mussa lacera (Pallas) Oken. 

 The calicles are often 40 to 60™°' broad, mostly isolated ; costae strongly spinose. 

 It is found throughout the West Indies, to South America, but not at the 

 Bermudas. See below, p. 130. 



