150 MADREPORARIA. 
In one of the type specimens of Ehrenberg the outer part of the wall of the radial 
corallites is not so thick as usual, and in some cases is quite thin, and in some specimens the 
branches are much more subdivided than in others. 
This and the following species are sometimes very difficult to separate and it may be that 
both are varieties of one species. M. pyramidalis is readily distinguished from M. seriata 
by the form of the branchlets, but there appears, so far as I can judge, to be no constant 
difference in the corallites. 
Indian Ocean: Red Sea, Seychelles, Mauritius, Ceylon, Mergui Archipelago, Great- 
Barrier Reef, ?Tongatabu. | 
a. Red Sea. Dr. Klunzinger [C.]. 86.10.5.42. (= M. pallida, Kz.) 
? db. Tongatabu. J. J. Lister [P.]. 91.3.6.15. (Young colony.) 
c. Low Woody Island. Sayille-Kent Coll. 92.6.8. 146. 
d. Capricorn Islands. Saville-Kent Coll. 92. 6. 8. 147. 
e. Troughton Island. H.M.S. ‘Penguin.’ 92.1. 16. 6. 
157. Madrepora pyramidalis. 
Madrepora pyramidalis, Klunzinger, Korallenth. d. roth. Meeres, Th. ii. p. 12, pl.i. fig. 2, ? pl. lv. fig. 6, 
pl. ix. fig. 7, pl. x. fig. B, pl. ii. fig. 3; Mobius, Beitr. z. Meeresfauna Mauritius, p. 45; Duncan, 
Journ. Linn. Soc. London, 1886, vol. xxi. p. 20; Ortmann, Zool. JB. 1888, Bd. iii. p. 151 (part.). 
Corallum low cespitose or incrusting. Branches pyramidal, rather crowded, so as to 
become angular below, but conical above, tapering rapidly to a blunt apex; branches little 
divided unless near the margin of a colony, and budding branchlets are not of frequent 
occurrence; branches 3 to 4 cm. high and 8 to 10 mm. diameter at a point 1°5 cm. from 
the apex. Radial corallites much appressed, rather unequal, chiefly with the mner part of 
the wall undeveloped. In other respects the species agrees closely with M. seriata, Erb. 
Var. depressa, Klunzinger, op. cit. pl. il. fig. 3. 
Corallum incrusting, without branches ; corallites all short, thick, tubular, those which 
represent axial corallites recognizable by their larger size. One of Klunzinger’s specimens in 
the Berlin Museum consists of a colony in which the branches at one side are of the usual 
type, but on the other gradually decrease in length until they reach the incrusting condition 
of var. depressa. The specimens referred by Bassett-Smith to this variety agree with the 
figure, but form large incrusting masses over an irregular surface of considerable extent. It 
is, of course, possible that they belong to M. smithi, with which they occur on the Tizard 
Bank ; probably in an incrusting condition the two species could not be separated. 
Var. corymbiformis. 
Corallum corymbose with a solid base. Branchlets on the upper surface of the charac- 
teristic form, but united together by the solid basal structures, 2 to 3:5 em. long and about 
