ACTINOSTROMA INTERTEXTUM. 139 



are a centimetre or less apart. Tbe astrorhizse do not appear to be arranged in 

 vertical groups, and the coenosteuin is, therefore, not traversed by vertical astro- 

 rhizal canals. The general surface is covered with exceedingly minute, close-set 

 tubercles, representing the upper ends of the radial pillars. 



As regards the internal structure of the ccenosteum, the " continuous " radial 

 pillars are very delicate (PI. XIII, fig. II), and, as growth is not effected by 

 means of " latilaminsB," they run apparently indefinite distances. About five pillars 

 occupy the space of 1 mm. The concentric lamina? are placed about as far 

 apart as the pillars, but instead of forming continuous lines, as seen in vertical 

 sections, they are incomplete, and give rise to a species of loose vesicular reticula- 

 tion, which is highly characteristic of the species. The connecting-processes, or 

 " arms," given out by the pillars, are numerous, slender, and very regularly deve- 

 loped, and they give rise, by their union, to a close " hexactinellid " network, the 

 meshes of which are mostly more or less triangular (PL XIII, fig. 10). 



Obs. — Actinostroma intertextum,} Nicli., is distinguished from most of the species 

 of Actinostroma by the general characters of its delicate radial pillars, the loosely 

 reticulate structure of the concentric laminae, and the fact that the ccenosteum 

 has the form of a thin laminar expansion, with a basal epitheca. The species with 

 which it is most nearly related is the Act ino stroma Schmidtii, Rosen sp., of the 

 Silurian Rocks of Oesel. As I have elsewhere pointed out, however ('Ann. Nat. 

 Hist.,' ser. 5, vol. xvii, p. 233), this latter species is distinguished by the fact that 

 its astrorhizas are of large size, with wide and very slightly subdivided branches, 

 and are arranged in vertical groups, each group being connected with a wide 

 vertical or axial canal. The network formed by the union of the " arms," as seen 

 in tangential sections, is formed of oblong or irregular, rather than angular 

 meshes ; and some of the radial pillars appear to be of decidedly larger size than 

 the others. Lastly, the ccenosteum appears to have been massive, though the 

 mode of growth is not perfectly known. 



The ordinary British examples of A. inter te.vtum, Nich., show no particular 

 variations from the general type. Examples from the Silurian Rocks of Esthonia 

 have the radial pillars decidedly more closely set, while the concentric laminae 

 are more completely developed, and are not of such a reticulate or vesicular 

 character. They also do not exhibit the same regular " hexactinellid" network in 

 tangential sections ; though this latter feature may be only the result of imperfect 

 preservation. I have therefore proposed (loc. cit. supra, p. 234) to indicate these 

 differences by giviug to the Russian examples the special name of A. mtertextum 

 var. suevicum. 



In the Silurian and Ordovician Rocks of Britain there occur specimens of wli.it 

 I am inclined to regard as examples of this species in a peculiar state of preserva- 

 tion. The specimens in question are exceedingly ill-preserved, and thin sections 



19 



