68 MADKEPORARIA. 



Tlie calicles are small, slightly more than 1 mm. in diameter, angular, never quite super- 

 ficial. The walls are sharp, thin, and straight, rarely zigzag. The septa project from the walls 

 and form septal granules. In older, shallower calicles the septal granules are near the wall, 

 but appear to he distinct from it as a second outer ring to the pali, here and there even rising 

 like the pali. The fossa is deep and obscure, and out of it the pali rise as long thick rods in 

 the older calicles, but in the younger they are small, ragged and irregular. They are mostly 

 six in number, the directives being sometimes distinguishable. The columellar tubercle is 

 small and slightly flattened. 



The colour of the unbleached stock is a yellowish-brown. 



This description is taken from my notes on, and a rough sketch of a specimen in the Paris 

 Museum (No. Z 182 h). PI. IX., fig. 1. is a photograph of the specimen, and shows that 

 its method of branching is different from that of any other known form. It has no resem- 

 blance to that figured by Dr. Vaughan* as P. Pontes forma divaricaia, whicli branches at 

 much smaller angles and seems to have more of an axial stem. 



In this coral the largest axial stem arising from the tangle appears. to break up early 

 into three almost symmetrically diverging branches, which gradually curve upwards, and each 

 at its tip seems as if dividing irregularly. This branching requires further investigation and 

 comparison with that of other specimens of the same coral found at the same spot. In the 

 meantime this account is hardly exact enough to be used as a definite case of triple forking. 

 The photograph does not support it, and the appearances can be explained more easily by 

 rapid and UTegular divisions of the tips into two. 



54. Porites Belize 2. (P. Belizei secunda.) (PI. I. fig. G ; PI. XVII. fig. 7.) 

 [Belize, coll. J. Smith ; British Museum.] 



Description. — The corallum is nearly hemispherical, the living layer extending down the 

 sides irregularly to 2, 3, or 4 cm. of the substratum. The surface is greatly roughened, not 

 only by the larger convolutions from 1-2 cm. across, which show a tendency to form a slight 

 ridge or keel, but also by smaller ridges roughening the larger and by sporadic risings of the 

 calicle walls. 



The calicles are of all sizes from 1 • 75 mm. in diameter on the ridges and convolutions to 

 • 5 mm. in the valleys. The larger have polygonal wall-ridges with open circular fossae ; the 

 smaller are sharply angular and crowded together, with only rudimentary intervening wall 

 frameworks. These rise up into sharp conspicuous lines, here smooth, continuous, and not 

 very zigzag, there broken up into rough, irregular denticulations. The wall ridges everywhere 

 tend to rise into sharp points in the angles and thus still fui-ther roughen the surface. 

 The septa commence as small, sharp pointed projections from these ridges, and slowly lengthen 

 as they descend in the calicle ; they are fairly symmetrically arranged and show but very faint 

 indications of meeting and fusing. The columellar tangle is so open and straggling that the 

 calicles on the uppermost part of the stock are the deepest I have found in the genus ; but 

 • Hull. U.S. Fish Commission, 1900, ii. p. 316, pi. ii. fig. 4. 



