158 MADREPORARIA. 



froDi Madagascar. The text does not make it absolutely certain that the specimen fifjured 

 itsell' came from that locality, for it is not quite safe to take the words exactly similar sjieci- 

 mens ("ganz iiliereinstimmende exemplare ")too literally. Differences which are now thouglit 

 to be of importance were then frequently not recognisable by persons unprepared for them. 

 Hence, there is an element of doubt as to the locality, and, consequently, as to my designation. 



The specimen has hitherto been accepted as a true Pontes, but owing to the fact that it 

 liail from fifteen to twenty .septa it has here been transferred to the genus Guniopora. 



If this is correct, tlie stock is of interest in the genus. There is no other known 

 representative of it with so fiueut a method of growth. Most are stiff and formal, see 

 I'lates XI. to XIV. of Vol. IV. In this connection it is worth noting that the calicles must 

 have been small, for in the account of figure 59«, wldch was apparently a true Poj'ites, its 

 calicles are said to have been larger than those of this coral. 



The confusion which has been caused by Esper's coupling of two such different corals as 

 those figured pi. lix. and UxA., as representatives of one and the same species, would fill several 

 jtages to relate. Brief refcirences will be found above, pp. 115, IIG, as to the way Lamarck was 

 infiueuced. Dana was the first to rename the branching form and call it Porites conferta. 

 Out of the confusion we emerge with the simple facts that there is a Goniopmi with this 

 remarkable branching growth-form, and that it may be found at Madagascar. 



ZANZIBAR. 



' There are three interesting remains of a Poritid, which appear to be those of Goniopora 

 from Tertiary (Neogene) formations of Zanzibar. 



a. The well-preserved remains of a Poritid showing a well-developed reticular skeleton 

 consisting of stout straight trabeculse joined by short thick cross-pieces, so that the meshes are 

 all uniformly small and rounded. There are no good transverse sections, and the portions of 

 such sections which are visible show no certain traces of the arrangements of the calicles. 

 Here and there one sees what appear to be rather large rings of interseptal loculi, but for 

 the most part, the meshes of the coral show no symmetry at all. The stoutness of the elements, 

 tnibeculiB and cross-junctions is a slight argument in favour of their being those of Goniopora, 

 and not of Porites. The specimen seems to have encrusted a finger-shaped mass of some coral 

 conglomerate, one end of which was swollen. The Goniopora had at one time enveloped this 

 swollen end, and the section siiows a clear sharp line between the earlier conglomerate and the 

 more recent coral. 



h and c are irregular, rough, bent, finger-shaped masses of the same coral conglomerate. 

 A cross-section of c shows it to have belonged to what appears to have been a similar I'oritid, 

 but apparently of a branching form, because tlie skeletal elements appear to have been arranged 

 concentrically. 



Notliing ccrtaincoM. be made out beyond the fact that these fragments are the rem.'uns of a 

 Poritid, and probably of a Goniopo7n. They are of great interest, and deserve more thorough 

 investigation. 



a, b, c. Presented by J. T. Last, E.sq. Geol. Dept. K. 5354. 



