Tocene Corals, Central New Guinea. 531 
p- 89; Hist. Nat. Cor., vol. 111, p. 187), but in that species the septa 
are more equal and the corallites are 3 mm. in diameter. 
P. belli, Greg. (1900, p. 223), from the Miocene or later rocks of 
Christmas Island, is an allied species, but has better developed walls. 
Monripora, Quoy & Gaimard, 1833. 
Montipora antiqua, u.sp. (Pl. XXII, Figs. 1a, b.) 
Diagnosis. —Corallum massive. Corallites long and thin. Calices 
circular and shallow; their diameter is about half the width of the 
distance between the calicinal centres. The calices are surrounded 
by a zone of porous tissue. The corallites are separated by short 
straight lines of more compact ccenenchyma, which gives the corallum 
in section the appearance of being composed of prismatic corallites. 
Septa twelve in number; they consist only of spines. Columella 
conspicuous. Pali absent. 
Figures.—Pl. XXII, Fig. la, photograph of a polished transverse section 
(No. 36); x 3diam. Fig. 10, polished vertical section of the same specimen ; 
x 3 diam. 
A finities.—This interesting coral has the structure well preserved 
on the surface which was first polished. A microscopic section was 
cut from the lower part, but the material there has been altered, and 
the intimate structure is less satisfactorily shown. The coral, 
owing to its spiny septa and massive porous ccenenchyma, belongs 
to the Montiporide. 
Among the living species of Montipora this species is nearest to 
those which, like Jf. "hirsuta, Bernard (1897, p. 164, pl. xxxiv, fig. 16), 
have a columella and the calices separated by a perforate wall that 
rises, as in J. calcarea (Bernard, 1897, p. 59, pl. xxxil, fig. 13), to 
a distinct ridge. If these two characteristics were always associated, 
as in I. antiqua, species in which they are present should be separated 
as a distinct genus or subgenus. ‘The recent species that have 
a columella occur in each of Bernard’s five sections of the genus; 
for among the glabrous group it occurs in JL exserta, Quelch ; in the 
glabro-foveolate in Jf. spatula, Bern. ; in the foveolate in IW. calearea, 
Bern. (though no columella is shown in Bernard’s figure of that species, 
1897, pl. xxxii, fig. 13); in the papillate in JL verrucosa, Lam., 
MM. venosa, Ehr. (not shown in Bernard’s figure, pl. xxxu, fig. 16), 
and M. spumosa, Lam.; and in the tuberculate in IL friatilis, 
Bern., J. ellisi, Bern., I. efflorescens, Bern., MW hirsuta, Bern., 
and I. incognita, Bern. Asthese five groups are based on so variable 
a feature as surface ornamentation they appear artificial, and the 
presence of a columella and of the immature wall in the coenenchyma 
are probably more constant and reliable characteristics. Milne 
Edwards and Haime in their diagnosis of Montipora say that it has 
neither columella nor pali.! But Bernard has shown that a columella 
is present in some recent Montipora, such as MW. ellisi and I. floreseens. 
Though it is inconsistent with the practice throughout the rest of the 
Madreporaria to leave in one genus species with and species without 
1 Duncan, by an obvious slip, wrote ‘‘ with columella and pali’’, whereas it 
should probably have been “‘ without columella and pali’’. 
