GENERA OF FAVOSITID.E. 85 



pora {Favosites) cervicornis. The presence of very minute 

 calices scattered among the larger ones is of itself, perhaps, 

 hardly of more than varietal value ; but if Goldfuss's figure can 

 be trusted {Calamopora polymorpka,\2ir. gracilis, Petref Germ., 

 PI. XXVII., fig. 5), then Pachypora {Favosites) dubia is suffi- 

 ciently distinguished from P. cervicornis by the comparatively 

 large size and the obliquity of its calices. 



Lastly, it seems very probable that some of the forms figured 

 by Rominger (Fossil Corals of Michigan) are really inseparable 

 from the present species ; and this is specially the case with F. 

 limitaris, Rom., in part or in whole. From an examination of 

 numerous specimens of this form from the Corniferous Lime- 

 stone of Ontario, I can assert its very close resemblance to ex- 

 amples of P. cervicornis from the Eifel — this resemblance in 

 some cases amounting to absolute identity of external char- 

 acters. Other specimens, again, have larger and more dis- 

 tinctly circular or annular calices, thus closely approximating to 

 the type of Pachypora {Favosites) dubia, De Blainv. Unfortu- 

 nately, the highly silicified state of the Canadian specimens, 

 accompanied with a more or less completely hollow condition 

 of the tubes of the corallites, renders it impossible to prepare 

 satisfactory microscopic sections ; and in the absence of these, 

 I hesitate to pronounce as to their absolute identity with 

 the European type, though I cannot doubt their very close 

 alliance. 



Taking the Eifel specimens as the type of Pachypora cervi- 

 cornis, De Blainv., the corallum is generally in the form of cy- 

 lindrical or subcylindrical, often irregularly swollen branches, 

 from two to ten lines in diameter (PI. IV., fig. 3), dividing at 

 variable intervals, but usually not anastomosing. Where inos- 

 culation of the branches occurs, it may perhaps be convenient 

 to use De Blainville's name of reticulata as a varietal designa- 

 tion, though forms so characterised certainly differ in no other 

 particular from those which divide without anastomosis. The 

 corallites radiate from the axis of the branch, curving gently 

 outwards till they open on the surface, and their true form is 



