PORITID^. 3 



dealt with. Apparently a mere question of technical method, it assumed an importance which 

 in itself it in no way deserves. 



In the Preface to the British Museum Catalogue of Fishes, Vol. VI., Dr. Giinther 

 observed that, owing to the extraordinary variability of the Salmonidae, the ordinary method 

 of distinguishing and determining their species was utterly inadequate. This is still 

 more true in the case of the Stony Corals. Indeed, I have no hesitation in asserting that 

 in the present state of our knowledge coral species are indeterminable. Nevertheless, 

 in previous volumes the specimens showing striking differences of structure were named 

 and called " species." This was done solely in accordance with long established practice. For 

 while, in the absence of the data necessary for their arrangement into genetic groups, there 

 was and could be no pretence that specimens arranged under the different headings were real 

 " species " in the strict sense of closely related groups, the names served a useful purpose as 

 nmemonics relating to the known structural variations presented by the genus. Experience 

 showed me, however, that this usefulness varies in inverse proportion with the size of the 

 group, that is, with the magnitude of the burden it has to undertake. The long lists of names 

 which were steadily growing as I proceeded to designate every apparently different form of 

 Porites in the old way, got completely out of hand. This fact, apart from the fundamental 

 confusion of idea involved in calling them "species,"* convinced me that if the systematic work 

 to which I had devoted so much time were to have any scientific value at all, some new method 

 of presentment must be adopted. The time ought to have passed when the value of a systematic 

 treatise is to be reckoned by the number of new " species " established. It must, in the future, 

 be sought in the accessibility of the details relating to new observations, in the accuracy of those 

 observations, and in the ability with wliich they are correlated with already ascertained facts. 



The method adopted will be described in detail in Sec. VII., and the iirst attempt at its 

 practical application will be found in the following pages of this volume. Its appearance may 

 be strange and forbidding for a time because it has not resulted in a " classification " in the 

 ordinary sense, wMch means the arrangement of the collection into so many representatives of 

 ideal species set up on the^slippery evidence of the, mostly, isolated, and often fragmentary 

 specimens themselves. But if there is no conventional attempt at a genetic classification, 

 because the facts do not justify any such attempt, there is, instead, a catalogue of the known 

 forms presented in such a way as to render the available facts readily accessible to students of 

 the genus. The data are so arranged that they may be supplemented at any time without the 

 necessity for any of the usual subversive re-shuffling and re-naming wliich at present makes 

 systematic work in a group like the corals a labour of Sisyphus. In time doubtless the 

 requisite knowledge will be available for the analysis which will reveal to us the species. For, 

 in the nature of things, the species can only be discovered by the same analytical processes 

 which enable us to separate out the genera.f 



* Of. on this subject an article bv Prof. Doderleiii, Zeitsch. f. Morphologie and Anthropologic, 

 iv. (1902) p. ;)94. 



t See further ' On the Unit of Classification for Systematic Biology, Pmc. Cambridge Philo- 

 sophical Society, xi. (1902) p. 268. 



