208 DE. G. J. HINDE ON SEPTASTR^A, d'oRBTGNY (1849), 



of small tubercles ; more frequently, however, the median line con- 

 sists of a shallow but distinctly impressed linear furrow (PI. IX. 

 figs. 1, 6). The general surface of the summit- walls and the upper 

 margins of the septa are likewise covered with minute blunted 

 tubercles. 



In certain portions of the surface of the corallum in some specimens 

 the corallite walls are much thinner and sharp-edged, indicating 

 that they had not reached the same stage of mature growth as the 

 thick-walled individuals. 



The bounding walls, or so-called theca, of the corallites are ori- 

 ginally extremely thin and delicate, and it is only by subsequent 

 deposition on their inner surfaces of calcareous material, to which 

 Lindstrom * has given the name of " stereoplasm/' that they become 

 thick and solid in the later stages of growth. The walls of adjacent 

 corallites are distinct f and separate from each other, though in close 

 contact ; at the immediate surface of the mature corallum the par- 

 tition is only indicated by the linear furrow between the calices. 

 The distinctness of the walls is most clearly shown in fractured 

 branches and stems, the division usually taking place between indi- 

 vidual corallites, each retaining its own wall. In transverse sec- 

 tions, also, each corallite exhibits a distinct wall. 



Septa. — In their early stages of growth the septa are thin, delicate 

 plates, of the same character as the bounding wall or theca ; in fact 

 they are formed by infoldings of the outer wall or, rather, in other 

 terms, as the septa are first in order of growth, the theca or wall is 

 merely the exterior continuation of the septal laminae (PL IX. fig. 8). 

 This point will be referred to more in detail when treating of the 

 minute structure of the septa. The septa are extremely variable in 

 their development. In some young corallites, as seen in sections of 

 branches of the type specimen, only four are present, in others there 

 are eight septa. In some nearty matured corallites. under apparently 

 abnormal conditions of growth, there may be only two septa, forming 

 a single partition across the centre of the corallite, in others there 

 are four or five, unequal in length. In normal mature corallites 

 twelve septa are shown in the interior, some or all of which may 

 extend to its central axis, where their free lateral margins unite, 

 and frequently are slightly intertwisted together (PI. IX. fig. 5). 

 IS'ot unfrequently some of the septa slightly curve about halfway 

 between the wall and the centre of the corallite, and unite by 

 their margins with the sides of adjoining septa (fig. 7). Two short 

 adjoining septa also occasionally unite, so as to form a closed loop 

 inside the theca, quite independent of the other septa in the same 

 corallite. I have not observed more than twelve septa in the lower 

 and central portions of any of the corallites ; the only indica- 

 tions of a third cycle of septa are slightly impressed vertical lines 



* " Forteckning pa svenska undersiluriska Koraller," p. 30. Ofvers.af Kongl. 

 SvenskaYeteBsk.-Akad. Forh, 1873. 



t Prof. Duncan states {I. c. p. 26) that " the union is so decided, that the 

 corallites are, and always were, inseparable," ■which is not the case in the type 

 or any other specimen that I have seen. 



