408 ZOOPHYTES. 



The Gemmiporidae assume glomerate and broad foliaceous forms. 

 The latter are often of large size, and consist of plates sometimes a foot 

 or more in width, either standing like broad vases on a pedicel, or 

 clustered and rolled around one another, resembling some large-leaved 

 plants. The whole interior of the plates or vases is covered with 

 polyp-flowers, and fresh buds are constantly appearing at the margin, 

 as explained in 66. The polyps have a broad convex disk, which 

 is fringed around with short tentacles in two or more series. 



The cells of the corallum are nearly cylindrical, and are evenly set 

 around with narrow and entire lamellae, appearing often like mere 

 strise of the inner surface. The calicles are thick with a granulous 

 exterior like the whole corallum, and the margin, though sometimes 

 subacute, is always stout. In a transverse section, the cells of the 

 corallum consist of a neat star of cellules radiating from a porous 

 centre. 



The Gemmiporidse are confined to the warm coral-reef seas. 



This family includes but two genera : 



Gemmipora. Calicles prominent, cylindrical or conico-cylindrical'. 



Astreqpora. Calicles obsolete or nearly so. 



GENUS I. GEMMIPORA. BLAINVILLE. 



Gemmiporidce foliacea aut glomeratce. Corolla porosa, super faie granu- 

 losa, caliculis cylindrids aut conico-cylindricis. 



Foliaceous or glomerate. Coralla porous, with a granulous surface 

 and cylindrical or conico-cylindrical calicles. 



The genus Gemmipora was established by Blainville from the Ex- 

 planaria of Lamarck, a mixed group united only on the ground of their 

 unifacial explanate form. The glomerate species were arranged by 

 Lamarck with the Astrseas. Ehrenberg, adopting the name Expla- 

 naria for the Orbicellse, with which he united one or two of the Gem- 

 miporce, placed the remainder of the Gemmiporse under Turbinaria, 

 a name which was proposed for this group by Oken, previous to the 

 introduction of Blainville's Gemmipora. The word is so near Tur- 

 binalia, one of Lamarck's genera, that Blainville's, which has been 



