CYCLOSTOMATA. 



617 



how far even this distinction can be relied upon. Forms of this second 

 type occur from the Ordovician rocks onwards. It would seem that 

 these two groups oiStomatopoi-c? should be generically separated from one 

 another, as the differences above pointed out are of substantial import- 



Fig. 45S. — A specimen 

 of Stomatopora dichotoma, 

 from the Great Oolite, of 

 the natural size. (After 

 Zittel.) 



\.-'.J- 



Fig. 459. — a, A fragment of shell with Stomato- 

 pora auloporoides and JT. inJJata growing on it, of 

 the natural size, from the Cincinnati group (Ordo- 

 vician) of America ; b, Stomatopora auloporoides, 

 enlarged ; c, Part of the same, enlarged still further ; 

 d, Stomatopora inflata, enlarged ; E, Fragment of 

 Hederella Canadensis, from the Devonian rocks of 

 Canada, of the natural size and enlarged. (Original.) 



ance. It would also seem certain that the Silurian and Devonian genera 

 Reptaria and Hederella (fig. 459, e), sometimes placed among the Corals, 

 should be referred to this family. In these doubtfully distinct genera the 

 colony is branched and adherent to foreign bodies by its lower surface, 

 the branches consisting of long, narrow tubular cells, which spring in an 

 alternating manner from one another, each branch being thus biserial. 



Allied to the preceding is the family of the Diastoporidce, in 

 which the polyzoary is usually encrusting and discoid or fan-shaped, 

 though sometimes foliaceous and erect. The tubular cells are in 



Fig. 460. — Mesenteripora {Bidiastopora) cervicorm's, natural size and enlarged. Jurassic. 



great part immersed, their ends only being usually free, and the 

 peripheral part of the colony may be formed of small, angular 

 " germinal " cells. The species of Diastopora, in a restricted sense, 

 form small crusts upon foreign bodies, and range from the Jurassic 

 to the present day. Allied types appear in the Silurian rocks, and 



