Study of Madreporarian Corals. 175 



is to endeavour to find out the lines along wliich tliey have travelled. 

 This can only be done by careful comparison of the available generic 

 and specific variations presented by each. I venture to think that 

 the importance of not a few structural features of the calyx would 

 be found to melt away before such a process. By way of an example 

 I may note that Milne-Edwards and Haime united the genus 

 Montipora with Pontes on account of the trabecular character of 

 the septa of the former. A more extended survey of the genus 

 Montipora would have shown that the variation in the structui'e of 

 the coenenchyma and of the anatomy of the calyx is very gi-eat; and 

 that the trabeculge which are so pronounced in a large number 

 of Montiporan types are purely secondary, and that every stage in 

 their development can be traced from the simpler types which ally 

 the genus with the Madreporidae. Again, Miss Ogilvie speaks of 

 the "importance" of certain structural features in the columella 

 of Turbinaria. Her type-method has led her to assume that what 

 she finds in T. mesenferina was necessarily of importance, I have 

 taken the trouble to look at about a dozen different types of Turbinaria, 

 and did not find a single other type showing Miss Ogilvie's 

 important feature. The columella in Turbinaria is astonishingly 

 variable. Again, in dealing with the same coral, Miss Ogilvie 

 assumes that the wall structures are formed nearly contemporaneously 

 with the septa, and thus the interseptal ioculi are typically cut off 

 round the rim from the canals of the coenosarc. A more extended 

 survey of Turbinarian types would have shown Miss Ogilvie that 

 the interseptal Ioculi very frequently run into the surface coenosarcal 

 canals. Or, again, Miss Ogilvie compares a calyx of Madrepora 

 and a calyx of Turbinaria, the calicles taken at random. Anyone 

 who has spent any time studying and comparing these genera 

 must be convinced that their highly specialized method of budding 

 has profoundly modified the daughter calicles. The axial calyx in 

 Madrepora and the skeleton of the single parent polyp of the 

 Turbinarian stock before the ring of daughter polyps have begun 

 to form the cup, are the only ones that could be relied upon for 

 structural comparison with other coral types. 



But if this doubt prevails as to the supreme taxonomic importance 

 of the grosser features of calicle structure, what are we to say of the 

 finer texture of the skeletal parts upon which Miss Ogilvie would 

 found a natural classification. A mere formative tissue, following, as 

 it must do, every form variation, however insignificant, is a shifting 

 sand upon which no one should try to build up a phylogeny. 

 Miss Ogilvie could only establish her case by showing that the 

 diiferent units of skeletal structure, fascicles, and trabecular had 

 some clearly recognizable morphological significance as such and 

 were not merely component parts, produced wherever required, 

 more or fewer as the case might be, of no more significance than are 

 the individual fibrils in a muscle or nerve fibres in a nerve strand. 

 A certain ordered complexity of structure is not enough : the 

 muscle fibril is an organized body of marvellous complexity, 

 yet it has no morphological significance whatever in settling the 



