80 BRITISH STROMATOPOROIDS. 



very close together, with few or no radial pillars in the interlaminar spaces ; while 

 those of the next group are further apart, and have their interlaminar spaces 

 crossed by short irregular pillars. Extremely well-developed astrorhizse are 

 present. 



This genus was founded by Dr. Murie and myself (loc. cit.) for the singular 

 8. columnare ( = Stromatopora Wortheni, Quenst.), first described by me from the 

 Devonian Rocks of North America (' Pal. of Ohio,' vol. ii, p. 253, Plate XXIV, 

 fig. 1). We included in the genus another form (viz. 8. retiforme, Nich. andMur.), 

 but this is in reality a species of Actinostroma, and is nearly related to A. verrucosum, 

 Goldf., sp. On the other hand, Stylodictyon columnare, Nich., is a very peculiar 

 type, and in the present state of our knowledge can hardly be referred to any other 

 genus. I am not, however, clear as to the position which the genus Stylodictyon 

 ought properly to occupy, as the characters of the type-species are in many respects 

 such as would give it an intermediate place between the Actinostromidce and the 

 Stromatoporidce. In vertical sections, the structure of the skeleton — apart from the 

 characteristic columns — conforms to that of the Actinostromidce, the concentric 

 laminge being very well developed, and the radial pillars not being obliterated. On 

 the other hand, tangential sections (Plate VII, fig. 10) do not show the cut ends of 

 the pillars, but rather show a reticulated tissue, similar to that of the Stromato- 

 poridce. 



Various Stromatoporoids show an approach to the structure of Stylodictyon 

 columnare as regards the peculiar vertical columns which intersect the entire 

 coenosteum. Stromatopora consors, Quenst., is an example in point. Much more 

 extended researches are, however, necessary before it can be asserted that the 

 structure of the forms in question is really the same as in the present genus, or 

 before we can deal more precisely with the type-species, 8. columnare. 



The columns of Stylodictyon may, perhaps, be compared with the large spines of 

 Hydractinia circumvestiens, S. V. Wood (Plate VI, figs. 11 and 12); but they 

 appear to be rendered quite solid centrally, by the complete obliteration of the 

 interlaminar spaces, and they do not, therefore, open by apertures on the surface. 



Fam. 2, Labechiid^, Nich. 



The coenosteum in this family is composed of large-sized calcareous vesicles, 

 which are usually lenticular in shape, but may be rectangular, and which are 

 arranged in superposed strata as regards either a basal plane or an axial tube. 

 The vesicles are traversed at intervals by " radial pillars " directed at right angles 

 to the plane of their strata ; or they carry the same structures in a rudimentary 



