9G MADREPORARIA. 



Group III.— ASTR.EOPORJ; GLOBULARES. 



Tlie calicles lengthen and increase in number in such a way that the canenchymatmis skeleton early rolls 

 over the edge of the epitheca, forming at first a hemispherical, and later, as new layers cover the old, 

 an almost completely globular mass. 



Species 11. Astrsopora punctifera. (PI. XXX. ; PI. XXXIII. fig. 18.) 



Astrma punctifera, Lamarck, Animaux sans Vert., ii. (1816) p. 260. 

 Jstrceopora pmnctifera, Blainville, Die. d. Sci. Nat., be. (1830) p. 348. 



Description. — Coralluin an almost smooth globe, the living layer growing round, closely- 

 applied to the previous growth ; the edge either arrested by a fold of epitheca or rolling over 

 the epitheca, which thus fails to arrest it. Wliere the epitheca forms folds over the advancing 

 edge, it covers up the calicles. The core of the globe consists of previous concentric growths, 

 some of which may have been detached and rolled over. 



Tlie calicles small and insignificant, immersed, valleys, however, occurring here and there. 

 The fully developed calicles are uniform in size, 2 mm., circular, with irregular margins, 

 crowded, being less than their diameter apart. Young calicles appear in all interstices wide 

 enough, as circular pits, lined with smooth membranous mural tissue with no traces of septa. 



In the adult calicles, six primary, very delicate septa are distinct, which almost meet in 

 the centre. Two other cycles are indicated, especially around the margins of certain calicles 

 which are separated by valleys, and tlms surrounded with a thick rampart of reticulum ; in 

 these, the costal elements of three cycles of septa radiate almost symmetrically outwards. 



The crenenchyma is a rather fine, solid reticulum, looking somewhat granular owing to 

 the short echinulse not ending in groups of fine minute spines, but rounded off at the ends. 

 Over the surface, while the costal element is distinct, the synapticular floors are only 

 represented by trabeculse. In the broken sections of earlier growths, the costal element is by 

 far the more important, although the synapticular element is arranged in perfectly regular 

 tiers. 



The lengthening calicles form tabulae. 



There is only one specimen of this coral, which bears a close general resemblance to the 

 largest specimen of A. listeri in the Collection. It differs, however, from this type both in 

 the characters of its calicles and of its coenenchyma, as well as in the manner of its growth. At 

 the same time, it is obvious that the pulvinate type of growth might easily pass into the 

 globular, and it is often not easy to say whether a particular specimen belongs to the one or to 

 the other. 



The specimen tallies well with Lamarck's description, although the habitat is different. 

 It further fully explains how Lamarck's specimen could be like a cannon-ball, without any 

 apparent point of attachment. It is worth noting that this point in Lamarck's description, 

 though repeated by Dana, was dropped by Milne-Edwards. The single specimen in the 

 National Collection, however, shows how liable these great spherical (and therefore top-heavy) 



