A. K Verrill — Bermudian and West Indian Reef Corals. 81 



they generally belong, the smallest being quite short. The larger 

 ones have wide but slightly marked paliform lobes and are broadly 

 rounded at the summits ; their surfaces are finely granulate, and 

 costulate close to the border ; their edges are very finely and regu- 

 larly denticulated. 



The under side is covered with elevated, lamellate, radial cost?©, 

 which are sharply and closely dentate on their edges, the teeth being 

 small and spiniform. 



Length of the coral, 75°^'" ; breadth, 60"^" ; width of the valleys 

 mostly 13 to 25'"™ ; depth, 10-18'"™. 



The type is from an unknown locality, but was supposed to be 

 West Indian. It belongs to the American Museum, New York. 



From its aflinities with Trachyphyllla, I think its origin is more 

 likely Indo-Pacific. 



There is a smaller worn specimen in the Museum of Yale 

 University, locality unknown. 



Subfamily Favitinee Yer., nom. nov. 



This subfamily is intended to include all the astreiform corals that 

 normally or chiefly increase by fission or by intracalicinal budding, 

 for these two methods intergrade completely and often coexist on 

 the same coral. It is thus nearly equivalent to Flssicella of Dana. 

 Paliform lobes or teeth are generally present. 



This group is very closely related to Mceandrince. The principal 

 difference consists in the more complete fission of the zooids and the 

 rapid and usualh^ complete isolation of the calieles, which may be 

 either circular or angular. 



Perhaps it would have been thought better by many to have con- 

 sidered the group a distinct famil}^ near MmandridcB, under the 

 name F'avitidoe. But the study of such species as Favia gravida 

 and F. fragum, in comparison with Maeandra conferta, M. Agassizii, 

 and M. clivosa, var. dispar, shows that the two groups nearly inter- 

 grade. 



The occasionally isolated calicles of M<xandra are structurally 

 identical with those of Favia. Perhaps the two groups are not 

 even of subfamily rank. 



I have used Favites as the typical genus from which to form the 

 family name, because the ultimate fate of Astrea and Favia is still 

 uncertain. (See p. 80.) 



