DESCKIPTIOiSrS OP SPECIES. . 71 



appear aiTanged in rows or curves sloping downward and outward from the 

 columella ; in other places they seem arranged in curves parallel to the septal 

 margin. But the granules have another arrangement across the curves, and 

 where each row of granules perpendicular to the margin (in the septal plane) 

 emerges at the surface is a corresponding dentation. Except near the 

 columella, the septal teeth are directed out^yard; near the columella, they 

 are directed inward. In a ground section parallel to the septal face the 

 calcification centers have one arrangement in nearly horizontal lines or 

 curves, evidently parallel to the septal margin, and another arrangement in 

 diverging lines perpendicular to the transverse lines or ciu-ves. The calcifi- 

 cation centers are nearer together in the lines jjarallel to the septal margin 

 than in those perpendicular to it. Summing up, the septa are composed of 

 ti'abeculse completely fused, with an area of divergence situated interior 

 to the wall, and have dentations corresponding to the points of emergence 

 of the trabeculse on the septal margin. It should be added that the septal 

 dentations are not acute, but are rounded. 



The origin and character of the columella have already been described. 



Wall. — The ground cross section of a corallum that I have studied is 

 not thoroughly satisfactory, but as the material at my disposal is limited, 

 this section must suffice. The septa, where they are rather remote, seem to 

 project through the wall, and between the distal ends a piece is inserted to 

 eff"ect the fusion. Apparently a true theca is present. But in other places, 

 where the septa stand near together, it appears that the outer ends of 

 the septa are enlarged sufficiently to join with each other directly and 

 form a pseudotheca. Apparently the above is the constitution of the wall, 

 but it can not be asserted positively that this is correct. The occurrence of 

 true and false theca alongside each other is not remarkable, as was pointed 

 out in the introductory chapter on the morphology of the coral skeleton 

 (13. 491). 



Aldriohia elegans sp. nov. ^ 



PI. IV, figs. 15 to 19. 



Corallum small, straight, or slightly curved in the plane of the longer 

 transverse axis of the calice; elongate; cross section, compressed elliptical; 

 attached by a small, short, nipple-like pedicel. Some specimens may ulti- 

 mately become free. The costfe are well developed, granular, rather broad, 

 those next the edges of the corallum often slightly broader than those on 



