RADIAL PILLARS AND CONCENTRIC LAMIN/E. 39 



minute vesicles and tubuli which are found in the skeleton-fibre of Stromatopora, 

 Stromatoporella, and Stachyodes, and thej are not so regular nor so continuous as 

 they are shown to be in Bargatzky's figures, while they have much thicker walls. 

 Moreover, if we examine thin sections of other specimens of what I believe to be the 

 same species (which is very probably the same as the Stromatopora capitata of Gold- 

 fuss) we find that the skeleton-fibre exhibits in thin tangential sections numerous 

 large, dark, rounded dots, which in longitudinal sections are seen to be really the cut 

 ends of dark rod-like bodies, the fibre itself being clear and transparent (Plate XI, 

 figs. 7 and 8). I take it, therefore, that in this case also we have really to deal 

 with a system of vertical canals, which run in the skeleton- fibre, and are connected 

 at intervals by cross branches, and that the diffei'ent appearances presented by 

 different specimens result from the infiltration of these canals in the one set of 

 examples with calcite, and in the other set with oxide of iron. 



(//) The Radial Pillars and Concentric Lamince. — If such a Stromatoporoid as 

 Actinostroma clathratum be examined, the skeleton is seen to consist of two 

 principal sets of structures, one " radial" or vertical, the other "concentric" or 

 horizontal. These may be termed respectively the "radial pillars" and the 

 " concentric lamina? " or " horizontal lamina? " (Plate I, figs. 9 and 12). These may 

 be exceedingly distinct, or they may be so far blended together as to be hardly or 

 not at all recognisable as separate structures, so that it is almost a matter of 

 necessity to deal with these two constituents of the skeleton in conjunction. 



In most Stromatoporoids the " concentric lamina? " are the most conspicuous 

 structures, as giving rise to the characteristic foliated structure of most of the 

 fossils of this group. The skeleton, in fact, will in most cases split more or less 

 readily in a direction parallel with these lamina?, and therefore tangential to the 

 general surface ; whereas it has little or no natural tendency to fracture in direc- 

 tions at right angles to the surface, i.e. parallel to the radial pillars. The lamina? 

 are never strictly "horizontal," but are more or less undulated or curved, the 

 entire skeleton being thus more or less obviously formed of concentrically disposed 

 layers. In certain forms, moreover, (e. <j. Actinostroma verrucosum, Goldf.) the 

 lamina? are only partially concentric as regards the general surface, but are 

 concentrically arranged with regard to a number of points or lines of growth. 



Successive lamina? are separated by interspaces which are usually much wider 

 than the lamina? themselves, and which are termed the " interlaminar spaces." 

 These spaces are most conspicuous in forms such as Actinostroma and Glathro- 

 dictyon (Plate I, figs. 9 and 1). Even in these cases the interlaminar spaces are 

 not absolutely continuous, but are intersected at right angles by the "radial 

 pillars," which spring from the lamina which bounds the interspace inferiorly and 

 extend upwards, sometimes falling short of the upper bounding lamina, sometimes 

 reaching it and sometimes being continued onwards through many successive 



