HICKSON— POLYTREMA 449 



these areas are smaller and more numerous than the foramina on the stalk. At the 

 extremity of an unbroken branch the outlines of the chambers may be even more 

 clearly traced. The terminal chambers seem to be inserted on the edge of the branch 

 like a series of biconvex or irregularly spherical bricks on the top of a chimney. The 

 outer wall is perforated by foramina, the inner wall turned towards the axis of the branch 

 shows two or three large spout-like apertures by which the protoplasm in the cavity of 

 the chamber communicated with the exterior (Plate 31, fig. 16 and Plate 32, fig. 24). At 

 the free edge of some of the chambers there is a cock's comb of short tubercles. The 

 axis of the branch sometimes terminates in a veritable forest of siliceous sponge spicules. 



The extent to which the outlines of the chambers can be seen on the branches 

 varies a good deal. In the "Providence" facies they can be seen only at the extremities 

 of the branches. In the "Amirante" and " Saya de Malha " facies they may be seen 

 along the whole length of the branches and in some cases on the upper part of the 

 stalk as well (cf. Plate 30, figs. 3, 4, 6, 7). 



When the internal structure of Sporadotrema is examined, several points of great 

 interest can be discovered. As in Homotrema it is only the outer walls of the chambers 

 that are perforated by foramina, the walls of the chambers turned towards the axis 

 are not perforated by foramina. In some of the larger branches and stems the axis 

 is occupied by almost solid skeleton around which the chambers are arranged, but usually 

 the axis exhibits a number of cavities and irregular spaces communicating with one 

 another by large apertures but bounded by very thick and solid* walls (Plate 32, fig. 20). 

 The size of the chambers varies a great deal with the size of the specimen examined 

 and varies also within wide limits in every individual. An average diameter for the 

 chamber of Sporadotrema cannot be given in this paper with any approach to mathe- 

 matical accuracy, but roughly speaking the average diameter of the chambers of the 

 "Providence" facies of Sporadotrema is ten times as great as that of the chambers 

 of Homotrema. 



Another very striking feature of Sporadotrema is the thickness of the outer wall 

 on the stem and larger branches. In some of the specimens from Providence the distance 

 between the outer wall of the chambers and the surface of the coral is no less than 2 mm. 

 (Plate 31, fig. 14). It is this thickness of the outer wall that is responsible in some 

 measure for the smoothness of the surface of the stem and main branches. At the 

 free extremities of the branches where the outer wall is thin the outlines of the chambers 

 can be readily distinguished on the surface, but on the older parts where the wall is 

 thicker all traces of these outlines are lost. 



The initial chambers of one megalospheric form of Sporadotrema have been discovered. 

 Three of these chambers can be seen in the section (Plate 32, fig. 29 ic.) distinguished from 

 the other chambers of the specimen by their regular shape and by the fact that, although 

 buried in the base of the stalk, their outer walls are clearly perforated by true foramina. 



The size of the central chamber is perhaps an important point in the general argument 

 that Sporadotrema is distinct from Polytrema. At first I was inclined to believe that 



* The term "solid" is used only in a relative sense, the whole skeleton is undoubtedly perforated by fine 

 canaliculi as is the skeleton of Polytrema according to the researches of Merkel. 



