406 BULLETIN 103, UNITED STATES NATIONAL MUSEUM. 



Columella variable in development, represented by an axial 

 papilla or by a distinct axial lamella. 



Other specimens from Station 6866, St. John, Antigua. — There 

 are in addition to those already described, from station 6866, 

 two large specimens and fragments representing three others. The 

 largest is 13 by 14 cm. in diameter and about 9 cm. tall. The calic- 

 ular and septal characters are similar to those of specimen No. 2 of^ 

 the foregoing descriptions. The primary septa in many calices are 

 decidedly thick, the thickness of the other septa decreasing according 

 to cycle. The columella, although it appears to be derived from the 

 septa, is an axial lamella and in many cycles is decidedly thick. 



The specimens described in the foregoing remarks are the ones that 

 have given me the most trouble in identification. They grade directly 

 into typical specimens, such as the one on which Duncan based his 

 original description, and those described on pages 403, 404 of the 

 present discussion. 



The following is Duncan's original description of Isasii^ea iurbinata: 



Corallum 7 inches high, subplane and irregularly convex above, broad and gibbous 

 at the sides, small and conical at the base, whence the corallites radiate; upper surface 

 ridged with the elevated margins of more or less polygonal, close calices. Corallites 

 very long, slender, and prismatic, excessively crowded. Walls united, simple 

 throughout. Calices very numerous, irregularly pentagonal, not deep, and not packed 

 geometrically. Margins existing as sharp ridges, not marked by the septa, but faintly 

 ragged; united, crowded, not deep. Septa small, not exsert, not arched, but slanting 

 irregularly downwards and inwards, except the primary, which stand up in the fossa, 

 and are easily seen. They are laminar, delicate, and crowded, slightly toothed near 

 the internal end, ragged above, and granular on the sides. The primary septa some- 

 times meet by their inner ends; the secondary and tertiary are subequal when there 

 are others. They are disposed in six systems. In fully developed calices there are 

 four cycles in four systems and three in the rest ; in other calices three cycles with an 

 occasional fourth order. The fourth cycle is very small. Septa straight, not crenulate, 

 but slightly ragged ; no external spines. Endotheca tolerably developed. From the 

 condition of the base, which has been rolled, no epitheca can be seen. Reproduction 

 by submarginal (close to the wall) gemmation. Diameter of the calices from 2 lines 

 to3J lines [4.2 to 7.3 mm.]. (From the Chert formation of Antigua Coll. Geol. Soc.) 



A specimen that agrees with Duncan's descriptions and figures is 

 represented by plate 100, figures 4, 4a. This coral puzzled me for 

 some time but it is almost typical Aniiguasirea cellulosa, in which 

 the intercorallite tissues have been mineralogically changed so as to 

 present the appearance of solid intercorallite walls ; however, in a few 

 places the calicular edges persist, showing separate calicular margins 

 between which is a lower intercorallite area crossed by thin costae. 

 The septal and columellar characters are precisely as in A. cellulosa. 

 I failed to find Duncan's type of Isastraea iurbinata in London, but 

 I am convinced that it is a specimen of Aniiguasirea cellulosa in which 

 the intercorallite tissues are solidified by secondary mineral changes. 



Localities and geologic occurrence. — -This is one of the commonest 

 corals in the Antigua formation of Antigua, where I collected and 



