122 A. K Verrill — Bermudlan and West Indian Repf Corals. 



crowded, but have rather wide interseptal spaces, in which are the 

 much thinner and narrower small septa ; by the usually deep, steep- 

 walled calicles ; and by the prominent, thin, lamelliform, rather dis- 

 tant, and only slightly serrulate external costae. 



4a. 

 4. 



Figure 4. — Isophyllia fragilis (D.). Portion of a specimen having many of the 

 calicles isolated, with the polyps partly contracted. Photographed from 

 nature. About -? natural size. 



Figure 4«. — The same, with the polyps. A specimen having two calk-lt-s isolated 

 and the rest in a long connected series. About § natural size. 



The collines maybe high, steep, and narrow, with a thin solid wall, 

 or they may be double-walled, with a groove on top ; or they may 

 be entirely disunited in some specimens, up to 2*5 inches (65 mm ) in 

 diameter. But these variations in the. collines may occur on a single 

 specimen. The septa are decidedly thinner, fewer, and much more 

 openly arranged than in dlpsacea of the same size, and the "latter 

 has shorter, much stouter, and more regular septa! teeth, and less 

 prominent, closer, thicker, and more spinulose costse. 



The original type* of Dana belongs to the Museum of Yaie 



* Vaughan (op. eit. , pp. 41, 42, 1901) erroneously refers this species to Colpo- 

 phyllia gyrosa. He says that from the descriptions "no specific distinction 

 between the two can be discovered." This statement seems absurd, for the 

 Mussa fragilis= Isophyllia was very well described and figured by Dana. He 

 also described and figured the strong, spiniform teeth of the septa. Such a 

 mistake seems unaccountable, and the more so because Quelch had already 

 referred it to Isophyllia, in addition to my previous determination of it. Tin- 

 type of I. fragilis, in the Yale Museum, I have now figured. (See pi. x\ i. 

 fig- 1.) 



