the Collection of the British Museum. 417 



tliin, sliovt, uprii^lit spines. Diameter of cells 1 to U luilllm., 

 their depth about G to 10 millims. 



Hob. Unrccordecl. P.. M. 



In tlie only specimen the outline fissure of the upper sur- 

 face is kidney-sliaped, the coraUum beinj^ deeply emarginate 

 where it had been iixed to the ground. The greatest dia- 

 meter is nearly 8 inches, the height 4 to 5 inches, the greatest 

 thickness 1 inch. 



This species differs from all its congeners in its mode of 

 growth, in the ample development of the epitheca, and in the 

 structure of the ca?nencliyma. The echinulation of the 

 surface is more delicate, and the cells are smaller and more 

 distant, than in cither of the other species (perhaps with the 

 exception of A. palifera^ which 1 have not seen) . 



The genus Astrivopora now comprises five species, three of 

 which were already Known to Lamarck; the fourth was de- 

 scribed and figured by Dana as A. pidvinaria (U.S. Expl. 

 Exped., Zooph. p. 415, pi. 29. fig. 3), and afterwards enunie- 

 rated as A. profunda by Verrill (in Dana, ' Corals and Cor. 

 Isl.,' Appendix). 



Astrea stelhdata of Lamarck (Hist. Anim. s. Vert. ii. p. 261) 

 and Gemmipora fungiformis of j\Iichelin (Mag. Zool. 1840, 

 Zooph. pi. 2) do not belong to this genus. The first is not 

 determinable ; and even if it should ])rove to be a distinct 

 species, it ought to be renamed, because Lamarck meant to 

 describe the totally different Madrepora stelluJata of Ellis and 

 Solander. Gemmipora fungtformis is one of the earliest 

 stages of Turhiaaria peltata ; the only difference which might 

 be pointed out from the description and figure is the extreme 

 porosity of the cfcnenchyma. But this condition is evidently 

 due to the mode of preparation, and is frequently found in a 

 similar degree in specimens of this and the other species of 

 Turbinaria. 



IT. Remarks on the Species of Seriatopora. 



1. Seriatopora lineata. B.^I. 



MiUepora lineata, Linnneus, 17o8 and 1707. 



MadrcjHira seriata, Pallas, 170(3; F.llis vt SolamU'r, pi. ;31. ih^a. 1,2. 



Seriatopora sitbiitata, Lamarck, 1810; M. Edwanls. 



The Millepora lineata of Linnosus is evidently the same as 

 the Seriatopora. subulata of ^lilne-Edwards (l)ut neither of 

 Ehrenbcrg nor of Dana). Linnaus's description answers ex- 

 ceedingly well to this si)ecies, and is even nuuh more to the 

 Eoint than Lamarck's unsatisfactory diagnosis. I'allas may 

 ave included several species under his Madrepora seriata ; 



