236 bulletin: museum of compaeative zoology. 



Corallum, compound, one of the Astrceidce agglomerates gemmantes accord- 

 ing to Duncan's classification. Reprodixtion by budding between the calices. 

 The form ramous, or in small heads (see Duncan's M. cyathiformis). Coral- 

 lites rather small, joined by costse. The surface of the extra corallite area 

 granulate, the granules placed along the costse, which become plainly visible 

 only in sections. Septa, in type species in three cycles, with very obscurely 

 dentate, almost entire margins, sides granulate. Pali before the first and sec- 

 ond cycles. The pali are small and delicate, and are not arranged in two 

 distinct crowns, although sometimes those before the second cycle are a little 

 nearer the wall than those of the first cycle. Endotheca present. Columella 

 usually of several stout pillars, each having a rounded knob-like upper surface. 

 The number of pillars varies from one to four. When there are one or two 

 large pillars a small one or two may often be seen. In section, the pillars are 

 joined by processes to each other and to the inner ends of the septa, so the 

 columella in sections appears spongy, but by careful examination the pillars 

 can be distinguished. 



This genus, although it groups with Orhicella and ColumnasfrcBa, is very 

 distinct. The surface between the corallite is very different from that of the 

 former, and a further distinction is in the columella. Orhicella has a false 

 spongy columella. In Munich, through the courtesy of Geheimerath Prof. 

 von Zittel, I had the opportunity of studying a fine suite of Columnasti'cea 

 striata (Goldfuss), the type species of the genus. ^ The columella in Colum- 

 naatrcea striata is a simple stjde. There is a single crown of six ])ali sur- 

 rounding it, standing before only one cycle of septa. In neither of these 

 essential generic characters does the Jamaican coral agree with Columnastrcea ; 

 the columella of the former, as already stated, is not a single style, but consists 

 usually of several pillars ; there are not six, but twelve pali. It seems impos- 

 sible to include the two in the same genus ; therefore I have suggested a new 

 generic designation for Duncan's Heliastrcea cyathiformis. 



In the following specific description there is some repetition of the descrip- 

 tions of characters given in the above. 



Multicolumnastraea cytthiformis (Duncan). 



Plate XXXVII. Figs. 5-7, and Plate III. Fig. 1. 



1865. Heliastrma exsculpta, Duncan (non Pteuss), Quart. Jour. Geol. Soc. 



London, Vol. XX 1. pp. 7, 8, 11. 

 1865. Heliastrcea cyathiformis, Duncan, Quart. Jour. Geol. Soc. London, Vol. 



XXI. pp. 7, 8, Plate I. Figs. 1 a, 1 h. 

 1868. Heliastrcea exsculpta, Duncan, Quart. Jour. Geol. Soc. London, Vol. 



XXIV. p. 24. 

 1868. Heliastra;a cyathiformi", Duncan, Quart. Jour. Geol. Soc. London, 



Vol. XXIV. p. 24. 



1 Edwards and Ilaime, Hist. Nat. des Coral, 1857, Tom. XL pp. 262, 263. 



