WEST INDIAN ISLANDS POKITES. 49 



their edges, here free, there creeping. The central regions rise up into a few tall peaks with 

 steep sides and joined together like those of a mountain range diminishing in size till near the 

 edges, where the surface becomes quite smooth. 



The calicles are large and conspicuous, and nearly uniform over the whole surface, about 

 1 ■ 5 mm. from wall-ridge to wall-ridge. Tlie walls are everywhere thick and reticular, and 

 differ chiefly in being here rounded and there with a sharp median ridge. The wall-network 

 seen sideways is light and open, but seen from above looks close and solid. The septa project 

 as smooth, short threads, of uniform size and thickness throughout, and along their lengths 

 individually. They show a perfect radial symmetry except where here and there two appear 

 as if they would fuse in the typical manner. The columellar tangle appears as a smooth, 

 compact or solid, circular floor in the fossa, surrounded by a symmetrical ring of deep, rounded, 

 interseptal loculi. 



The colour of the unbleached coral is a deep brown. 



This is one of the massive forms usually called " astrceoides," but it is impossible to ignore 

 the difference in its calicles from those of the usual type shown in PI. II. fig. 2. A comparison 

 between this and such forms as the two described below and figured as P. West Indies x 27 

 and ^8, shows clearly tliat the old " astrceoides " is by no means a homogeneous group, see 

 Introduction, p. 15, and for the limitation here adopted of the term " astraeoid," see Table IV. 

 p. 142. 



a. Zool. Dept. 99. 6. 26. 6. 



31. Porites Barbuda 1. (P. Barhudce. prima.) (PI. II. fig. 8 ; PI. X. fig. 3.) 

 [Barbuda, coll. Gregory ; British Museum.] 



Description. — The corallum is branching, rising apparently from a stout, nodular base, 

 from which short thick processes project in different directions and either remain as mam- 

 millate stumps or flatten and begin to fork. If they succeed in forking, one prong in each 

 case usually remains aborted, sometimes at so early a stage of its growth as to make only a 

 knee-bend in the stem. The forkings, or attempts at forking, seem to take place at about 

 6 to 8 mm. apart. The growing branches are about 1 • 5 cm. thick. The living layer is about 

 5 cm. deep. 



The calicles, shallow and subcircular, vary from 1 • 5 to 2 mm. in diameter. The walls 

 are thin but without definite order of the skeletal elements and therefore very irregular. This 

 is partly due to the irregular thicknesses of the skeletal elements, which are everywhere 

 smooth and glistening. They are here thin and filamentous, there broad and flaky, else- 

 where again as stout granules ; in all parts these skeletal elements run into one another as if 

 melting together without any sharp angles or regular pattern. The zigzag of the wall-thread 

 appears chiefly in the arrangement of the smooth, swollen granules, which represent the points 

 where the septa abut upon the walls. The septa project very variably and irregularly from the 

 walls, but continually meet the columellar tangle, though the formula is somewhat obscure, 

 the septa being frequently bent about in such a way as to fill the calicle up with an open 



H 



