PA L^ONTOLOG V. 89 



These erect tubes are circular, closely approximated, or sometimes 

 in intimate, mutual contiguity, pressing each other into a sub- 

 polygonal shape. The orifices exhibit a cycle of longitudinal fur- 

 rows and intermediate rows of spinulose projections in rudimentary 

 development. Diaphragms are generally not developed. Occurs 

 in the Hamilton group at Stony Point, Thunder Bay. 

 Plate XXXIII. — Fig. 4, upper specimen. 



COLUMNAR!^. 



Consisting of the single genus Columnaria. 

 COLUMNARIA, Goldfuss. 



Goldfuss at first included within this genus several corals which 

 have a different structure ; he acknowledges, in the appendix^to 

 his work, however, that the coral called by him Columnaria sulcata 

 is only a weathered specimen of Cyathophyllum quadrigeminum. 

 Columnaria laevis is also, in all probability, generically different from 

 his Columnaria alveolaris, which alone is at present considered the 

 typical representative of the genus. In Goldfuss's characteristics is 

 expressly stated the absence of transverse diaphragms in the tubes, 

 although in his figures the diaphragms of the tubes are so distinctly 

 delineated that I can not conceive how they escaped his observa- 

 tion. I define the genus as follows : 



Convex colonies of contiguous, polygonal, or rarely of free circu- 

 lar tubes, growing from a few attached mother-tubes by rapidly 

 multiplying lateral gemmation. Tubes radiated by vertical lamellae 

 of alternately larger and smaller size, the larger ones in some 

 species reaching to the centre, or not. Number of lamellae from 

 twenty to forty. 



Transverse diaphragms simple, flat, moderately close in position. 

 Walls not perforated by pores, thin, and inseparably united in the 

 forms with polygonal tubes, appearing to be formed of simple lam- 

 inae dividing the adjoining cavities, but in favorably preserved speci- 

 mens the duplicity of the walls is positively observed. 



