I30 THE GENUS MONTICULIPORA. 



microscope shows the existence of a moderate, but not exces- 

 sive, number of the minute openings of the small corallites 

 (PI. V. fig-, i^). The clusters of the larger tubes are also not 

 conspicuously elevated above the general level. 



Tangential sections (fig. 21, a and b) show that the corallites 

 in the outer portion of their course have much-thickened walls, 

 their visceral chambers becoming rounded or oval. Their walls, 

 also, become seemingly completely fused in this region, so that 

 the original lines of demarcation between adjacent tubes can- 

 not be recognised at all. In the thickened walls between the 

 larger tubes are seen the openings of a considerable number 

 of interstitial tubules, which vary much in size, and are mostly 

 circular or oval. 



In the axis of the branches, transverse sections (fig. 21, d) 

 show that the corallites are at first quite thin-walled and poly- 

 gonal in shape, the same thing being shown in the central 

 parts of longitudinal sections. These latter (fig. 21, c) also 

 show that the thickenino- of the walls of the corallites beoins 

 almost as soon as they commence to bend outwards to the 

 surface, while a marked difference in the tabulation of the 

 larger and smaller corallites is displayed, — the latter having 

 much more numerous tabulae than the former. In all the 

 corallites, however, the tabulae are well developed, and they 

 are always complete and approximately horizontal. 



A consideration of the preceding characters will show that 

 I\I. Andreivsii, Nich., has no real affinity to M. pulchella, E. 

 and H. This is sufficiently proved by the complete amalgama- 

 tion of the walls of the corallites in the circumferential region 

 of the corallum, and by the presence of a set of closely tabu- 

 late interstitial tubes intermingled with the ordinary corallites. 

 On the other hand, yJ/. piilchella, E. and H., has no series 

 of interstitial corallites, and the original lines of demarcation 

 between the tubes remain permanently recognisable. 



The species to which AI. Aiidrcivsii is perhaps most nearly 

 allied is ]\1. rainosa, D'Orb. ; but the two are readily distin- 

 guished not only by the marked difference in their external 



