30 Prof. J. W. Gregory on the 



B. prolifera. 



3. No. 46. 7. 30. 8. Wreck Bay, Groat Barrier Reef, 

 N.E. Australia. 



Tliis specimen is the only one in the collection which is 

 catalogued as a Pacific form of prolifera. It is a small 

 fragment, 2 inches long, and its evidence is insufficient. 



Form intermediate between prolifera and cervicornis. 



4. No. 93. 4. 7. 43. Tahiti. (J/, regalis, Ehr.) 

 IMilne-Edwards & Haime * described M. regalis as " tr^s- 



voisine du J/, prolifera^ mais ayant les branches plus grosses." 

 That definition accurately describes the habit of this specimen. 

 The thickness of the branches agrees witli that of M. cervi- 

 cornis. The specimen differs from M. cervicornis by having 

 very short (1-2 millim. exsert) terminal corallites, in which 

 the primary septa are very unequal. According to Brook the 

 axial corallites of cervicornis have the terminal corallite 

 6-8 millim. exsert and the primary septa subequal. 



The corallites diff'er from those of M. prolifera by having 

 well-developed septa, whereas, according to Brook, in that 

 form " the directive septa are moderately developed, but the 

 remaining members of the primary cycle are more or less 

 rudimentary." 



C. cervicornis. 



5-7. Nos. 92. 6. 8. 210-212. Port Darwin. Saville Kent 

 coll. 



These three specimens, though differing somewhat in the 

 relative closeness of the branches probably belong to the same 

 species. The habit is cervicorn ; but the most striking feature 

 of the coral is that the terminal axial corallites are broad, 

 short, and thick-walled ; the wall, in fact, is equal in thick- 

 ness to the transverse diameter of the calice. In the youngest 

 corallites there are 6 septa ; in older corallites the septa 

 number from 12-16 ; in the largest and best-jH'eserved coral- 

 lites {e. g. in one marked with an ink-dot on specimen 

 92. 6. 8. 210) there are 3 complete cycles of septa. 



Now, according to Brook, in the subgenus Eumadrepora the 

 axial corallites have a " relatively thin wall and 12 septa." 

 Hence the Port Darwin corals are not typical members of 

 the same subgenus as M. cervicornis. 



* Hist. nat. Cor. vol. iii. p. 1-j9. 



