DEVONIAN ROCKS OF WESTERN CANADA. 39 



Billings rightly regards as not of generic value. Thus, Emmonsia 

 is distinguished from Favosites simply by the incompleteness of the 

 tabulie ; but this same peculiarity can sometimes be observed even 

 in F. Gothlandica, the type-species of Favosites, individual examples 

 of which not very uncommonly exhibit portions with the complete 

 tabulse of Favosites, and other portions with the imperfect tabulge 

 of Emmonsia. In the same way, there are many examples of 

 Favosites, of more than one species, in which the tabulae, whether 

 naturally or from some peculiarity in the manner in which they were 

 preserved, are quite rudimentary, and are even more imperfect than 

 they are in ordinary specimens of Favosites hemisjjherica, upon which 

 the genus Emmonsia was founded. 



The genus Astroeerium, again, was founded by Hall (Pal. IST. Y. 

 Vol. ii. p. 126) to include corals exceedingly like Favosites in all 

 superficial characters, but diifering in the possession of spiniform 

 septa. It would appear, also, though this character is not specially 

 mentioned, that Hall believed the corals which he referred to Astro- 

 ceriuin to be destitute of mural pores. This latter point, if it could 

 be proved, would amply suffice to separate Astroeerium generically 

 from Favosites; but there is much reason to think" that the apparent 

 absence of mural pores may be due to mineralisation merely. At any 

 rate, it is far from uncommon to meet with examples of undoubted 

 species of Favosites in which no pores can be detected. The other 

 point — namely, the presence of spiniform septa — is also not a satis- 

 factory distinction, partly because some examples of Favosites exhibit 

 the same thing, and partly because some examples which would 

 generally be referred to Astroeerium from theii' geological position 

 and general appearance, are Avithout any traces of septa. Upon the 

 whole, therefore, it would appear that the genera Astroeerium and 

 Emmonsia cannot be retained. 



The numerous species of Favosites may be divided into two groups, 

 according as they are massive or ramose, F. Gothlandica being the 

 type of the former, and F. polymorpha of the latter. The characters 

 wluch have been relied on as separating the species of this genus are 

 chiefly the following:—]. The diameter of the corallites ; 2. The 

 equality or inequality in size presented by the corallites; 3. The 

 completeness or incompleteness of the tabulae; 4. The number of 

 rows of mural pores ; 5. The position of the mural pores, whether 

 on the flat faces of the polygonal corallites or on their angles ; 6. 



