396 Dr. H. A. Nicholson and Mr. A. H. Foord on a 



Our observations have extended over a very large series of 

 specimens, and have led us to conclude that the size of the 

 tubes is an exceedingly variable character. That the tubes 

 of R. stromatopo7'oides, Roem., are on an average decidedly 

 smaller than those of B. crmah's, Schliit., is undeniable ; and 

 hence this character is one very serviceable in the discrimina- 

 tion of specimens of these two forms. On the other hand, 

 there are individual specimens, not otherwise separable from 

 the type of R. crinalis, which have tubes considerably smaller 

 or larger than the average of the species. In the same way 

 there are individuals of R. stromatoporoides^ E,oem,, with tubes 

 decidedly larger than is usual in the species, while others 

 have tubes below the ordinary standard of width. Hence 

 there are specimens which it is difficult to definitely refer to 

 either the one species or the other. Moreover, we find that the 

 size of the corallites is not necessarily or always constant even 

 in a single individual. In the case of R. stromatoporoides, in 

 particular, we find that a single specimen, or even a single 

 slide, may show in different parts precisely the same varia- 

 tions in the sizes of the tubes which Professor Schlliter relies 

 upon for separating his Calamopora piliformis from R. stro- 

 matoporoides^ Roera. We are therefore of opinion that, 

 except within certain restricted limits, the dimensions of the 

 corallites in these corals cannot be safely trusted to as a means 

 of discriminating species. 



The most interesting feature in connexion w^ith R. stroma- 

 toporoides, Roem., is, however, the extraordinary variations 

 exhibited by different individuals of the species as to the con- 

 dition of the visceral cavities and walls of the corallites. 

 These variations form a connected series, of which the follow- 

 ing are the two extreme terms : — 



(A.) In one set of specimens the visceral chambers of the 

 corallites are filled only with clear calcite, and the walls of 

 the tubes remain perfectly distinct. Such specimens also 

 have the tabulse and septal spines weU developed (PI. XV. 

 figs. 5-5 h, 6, 6 a, 7, 7 a). These examples closely resemble 

 R. crinalis, Schliit., in their main structural features ; but 

 their tubes are on the average decidedly smaller than those of 

 the latter species. Thus the corallites of R. crinalis have an 

 average diameter of ^ millim,, whereas the corallites of the 

 forms here under consideration are between j and \ millim. 

 in diameter. 



Judging from the short description given, we should say 

 that it is upon specimens of this group that Professor Schliiter 

 has based his Calamopora piliformis {loc. cit, supra) , and we 

 may therefore provisionally speak of such as "■piliformis " 



