CLASSIFICATION. 23 



ities of the Halysitidce, as here restricted. Upon the whole, 

 I am inclined to think that Halysites will find its nearest ally 

 in the Helioporidce, which it resembles in one important fact 

 namely, in the general possession of two sets of tubes (indi- 

 cating the existence of two sets of zooids). Moreover, the tubes 

 of the one series are of large size, with remote tabulae ; while 

 those of the other set are small, and have close-set or vesicular 

 tabulae. If this conjecture be correct, then Halysites must be 

 removed from the Zoantharia to the Alcyonaria (a course 

 which several authorities have already followed) ; but it will in 

 any case form the type of a distinct group. 



Fig. 10. A, Fragment of the corallum of Tetradium minus, Safford, from the Lower Silu- 

 rian of North America, of the natural size ; B, Transverse section of the same, enlarged 

 ten times, showing the petaloid form of the tubes and the short septa ; c, Vertical section 

 of the same similarly enlarged, showing the tabulse. 



VIII. TETRADIID^;. This group includes only the singular 

 Silurian genus Tetradium, Dana, in which the corallum is 

 massive, and is composed of polygonal or subcylindrical coral- 

 lites of great length, and closely contiguous. No "mural 

 pores" or other openings in the walls are present. The 

 tabulae are numerous and complete, and the septa are few in 

 number, typically four, short, and seeming as if formed by 

 inflections of the wall (fig. 10). Most of the examples of 

 Tetradium have the form and general appearance of Ch&tetes 

 radians, Fischer, and its allies; but Safford (Amer. Journ. 

 Sci. and Arts, ser. 2, vol. xxii. p. 236, 1856) records the 

 occurrence of specimens in which the corallites are united in 



