Structure of Heteropora neozelanica, Busk. 421 



these. In neither genus are we acquainted with the soft parts ; 

 and therefore we cannot assert positively that this dimorphism 

 has precisely the same significance in the two genera, while 

 there are grounds for thinking that the reverse is the case. 



(3) As regards the structure of the watt, the visceral 

 cavities of the tubes of Monticulipora appear to be always 

 closed, no traces of any pores or canals in the wall having yet 

 beeu clearly proved to exist. In the case of Heteropora, on 

 the other hand, the thickened walls of the tubes, in the peri- 

 pheral part of their course, arc traversed by an exceedingly 

 well-developed series of transverse canaliculi, which open into 

 the cavities of the tubes by definite pores, and which thus 

 place the body-chambers of contiguous zooids in direct com- 

 munication. These " canaliculi " differ structurally from the 

 " mural pores" of the Favositidae chiefly in being tubes with 

 definite walls and dilated extremities, instead of being mere 

 circumscribed deficiencies in the wall. 



(4) No radiating u septa," in the form either of spines or 

 of lamellae, are known to exist in any species of Monticuli- 

 pora. In Heteropora, on the other hand, the tubes, in the 

 peripheral part of their course, are intersected by numerous 

 delicate spinules, which are arranged in a radiating manner, 

 and reach a considerable distance into the body-chamber 

 (sometimes nearly to its centre). The spinules in form and 

 arrangement precisely resemble the " septal spines " of many 

 species of Favosites; but, admitting the Poly r zoan affinities 

 of Heteropora, it is obvious that they cannot be compared 

 homologically to the septa of any Ccelenterate. 



(5) Transverse partitions, or " tabulae," are universally 

 developed in the corallites of Monticulipora; and it is very 

 common for the different kinds of tubes which make up the 

 corallum to show marked differences in the nature and degree 

 of their tabulation. In Heteropora neozelanica, Busk, tabulae 

 are, so far as I have seen, always present, though their 

 number is comparatively small. They are also undoubtedly 

 present in other species, and in greater numbers (e. g. in H. 

 conifera, Haime, and H. pustulosa, Haime). So far as H. 

 neozelanica is concerned, the tabulae seem to be confined to 

 the axial region of the corallites, and not to be developed in 

 the interstitial tubes at all, thus differing in both of these 

 respects from the tabulae of Monticulipora. As in the case of 

 the radiating spines, however, just noticed, if we concede the 

 Polyzoan affinities of Heteropora, then the transverse parti- 

 tions which cross its tubes must have a different value and 

 import from the " tabulae" of Favosites and of the so-called 

 " Tabulate Corals " in general. 



