DESCKIPTIOMS OF SPECIES. 81 



the septa of the first cycle' trifurcate a short distance above the base. This 

 method of trifurcation is shown in PI. V, fig. 20. Septa in three cycles, 

 24 in number. Those of the first two cycles reach the columella. Their 

 surfaces granulate. Columella lamelliform. Seen from the side it presents 

 an outline like the base of a cordate leaf. The central projection seen from 

 above shows that it consists of two elongate little knobs, one on each side of 

 the vertical axis, i. e., along the shorter trausve'-se axis of the calice. These 

 little knobs are shown in the upper sui-face view of the calice, PI. V, fig. 20a. 



Mm. 



Greater tra,nsverse axis of calice 1-8 



Lesser transverse axis of calice 1 



Height of corallum 2. 7 



Locality. — Claibome, Alabama. 



Geologic occurrence. Claibome Sauds. 



PI. V, figs. 20 and 20a, are drawn from Lea's original type. Lea's 

 figure is not absolutely correct in the details of the costaj. In the side 

 view (PI. V, fig. 20) the specimen was turned a little. Lea's type is a 

 young specimen. 



The following additional notes are based on material in the United 

 States National Museum. 



The measurements for the large specimen (PI. V, fig. 22) are: 



Mm. 



Greater transverse axis of calice -' 5 (or more). 



Lesser transverse axis of calice 3 



Height of corallum 7.3 



The costa are rounded in profile and are granular. 



This specimen might be Sp. claibornensis sp. nov. Its state of preser- 

 vation is not perfect. 



In order to get an accurate comprehension of the species, it seems 

 best to trace its development as closely as the material at my disposal will 



permit. 



The specimen which shows the youngest stage of all the specimens 

 tliat I possess is 2 mm. high. It shows four costse originating at the bottom 

 of the base; two other costse, one standing at each end of the longer trans- 

 verse axis, are introduced very soon after, making six. One more costa is 



' In this description the cycles are ilenomlnated as traditionally understood. 

 MON XXXIX 



