﻿TUBERCULATE MONTIPOE^. 135 



without the convex basal plate occur, which are merely creeping down still younger branched 

 forms. 



The calicles are star-like and minute (0 • 75 mm.), very clearly and sharply cut out of the 

 smooth, flake-like lamella of the ccenenchyma ; the aperture level with the nearly smooth 

 surface and not closely surrounded by tubercles. The primaries reach to the half radius circle ; 

 secondaries merely notch the neat, round interseptal loculi between the primaries, making the 

 latter heart-shaped. The calicles on the under surface minute (0*5 mm.), numerous, star- 

 shaped, also cut cleanly out of the thicker and coarser lamellse of the coenenchyma. 



The ccenenchyma in section of an explanate portion shows a laminate streaming layer, 

 the lamiuEe having circular holes ; the upper thickening layer is very markedly trabecular, 

 the trabeculse being long and distinct, joined together by horizontal flakes in such a way 

 that the spaces between them are circular in vertical section. The trabeculse shoot up above 

 the surface into tall tubercles, often thin and flat, with small circular holes through them. 

 These, often bristly looking, tubercles are sometimes unevenly distributed, but are always 

 especially crowded and long on the rising knobs, or on the tips of branches. On the broad 

 flat tops of fused branches they may unite to form broken systems of tall ramparts round the 

 calicles. They usually stand back a little from the calicle margin. 



The surface of the interstices between the tubercles, being composed of lamellse per- 

 forated by a few circular holes, is smooth, and where the pores are not numerous, stony- 

 looking. 



This species, which is extraordinarily variable, is represented in the National Collection by 

 six specimens, three of which show the typical method of growth with the large convex 

 basal plate. Dana originally gave the name M. hispida to the specimens with this plate, 

 while those without it, though from the same locality (Singapore), he called M. " spumosa." * 

 Milne-Edwards and Haime first pointed out the probable identity of the two types. An 

 examination and comparison of the specimens, compared with what we know of the habits of 

 the genus, suggests a possible explanation of the two forms. The young coral commences by 

 encrusting the tips of the branched corals down which it may creep, successive growths 

 adding to the thickness of the stock. Under favourable circumstances the lowest fringe of 

 the creeping edge may expand and be successively widened and thickened (one large specimen 

 shows traces of seven encrustations forming the basal disc), until a large, circular, crinoHne- 

 like growth depends on all sides, from the top of which the stock proliferates in lobed 

 branches. Stocks thus get heavier and heavier, but as the crinoline expands it finds new 

 supports in other corals. The largest specimen in the collection is 40 cm. in diameter and 

 20 cm. high, and seems to have rested on a few corroded branches now deeply imbedded in 

 its substance and almost rotten from the action of boring organisms ; the under surface either 

 forms pendent drops, or else patches of the coenenchyma appear protruding through the 

 epitheca, probably for fresh points of attachment. 



a and b are two large specimens showing the typical development. They are probably 



* See Dana's text : under the figure the name, evidently by a printer's error, is " Squamosa" 



