422 On the Structure of Heteropora neozelanica, Bush. 



(6) Lastly, as to the supposed relationship between Mon- 

 ticulipora and Heteropora, and as to the deduction which has 

 been drawn from this as to the propriety of referring the 

 former genus to the Polyzoa, it is clear that the points of like- 

 ness between the two are by no means so weighty as the 

 points of difference. On the one hand, we have a strong ex- 

 ternal resemblance, a general similarity in the mode of con- 

 struction of the skeleton, and an agreement in the fact that 

 in both genera the colony consists of two sets of tubes, 

 while both have their tubes crossed by transverse partitions. 

 Such transverse partitions of the tubes (or, as we may loosely 

 call them, " tabulse ") occur, however, in organisms of such 

 exceedingly diverse affinities that we can, admittedly, attach 

 no value to the last mentioned of the above resemblances. A 

 mere similarity in general form, appearance, or mode of 

 skeletal conformation is also of no classificatory weight, 

 since we could find species of Favoaites or Pachypora which 

 in these respects are quite like either Monticulipora or Hete- 

 ropora ; so that, after all, the resemblances between the two 

 genera under consideration dwindle down to a comparatively 

 small quantity. 



On the other hand, to set against the mostly superficial 

 points of resemblance above noted, we have a number of 

 fundamental structural differences. Thus, in Monticulipora 

 the walls of the tubes are imperforate, there are no traces 

 of radiating spines or " septa," and in the dimorphic or 

 trimorphic species there are usually important structural 

 differences as regards the different groups of corallites. In 

 Heteropora, on the contrary, the walls of the tubes are 

 traversed by a very remarkable and exceptionally developed 

 canal-system, the tubes possess in their outer portions a 

 well-developed series of radiating spines arranged in vertical 

 rows (sometimes, at any rate, if not always*), and the 

 interstitial tubes are in no way structurally different from the 

 proper zocecia. 



In the face of the above distinctions I feel compelled to 

 believe, in the meanwhile, that there is no real relationship 

 at all between Heteropora and Monticulipora. This belief 

 would not, of course, constitute any valid ground for denying 

 the possibility that Monticulipora may truly belong to the 



* It is true that radiating spines have not generally been observed in 

 Heteropora, and that even in H. neozelanica, where they are plentifully 

 developed, they seem to have been overlooked by such excellent ob- 

 servers as Prof. Busk and Mr. Waters. I ascribe this, however, to 

 their very fragile nature, and to the general neglect of tangential sec- 

 tions, in which alone they can be readily made out ; and I entertain no 

 doubt that they occur generally in the genus. 



