376 



SCIENCE 



[N. S. Vol. XXVI. No. 664 



group; some even doubt wliether it can be 

 divided into species as are the mammals, birds, 

 mollusks, etc. The solitary Madreporaria are 

 probably not very much more difficult to study 

 than other lowly organisms that secrete exo- 

 ekeletons; but it is different with the com- 

 pound forms. The difficulty in the way of 

 making adequate studies is increased by the 

 usual insufficiency of material, which has al- 

 most never been collected in such a way as to 

 furnish data on the physical environment 

 under which it lived, and the entire absence 

 of any information obtained through experi- 

 ment. In working over collections as they are 

 usually submitted for determination and re- 

 port, one is often entirely at a loss whether to 

 refer certain specimens to previously described 

 species, or to consider them new. As the 

 specimens may not exactly fit any hitherto 

 characterized species, there is, naturally, hesi- 

 tancy to apply an already established name; 

 should they be described as new, subsequent 

 collections may render the name applied in- 

 valid. As matters now stand, we know a 

 large number of characters of the Madrepo- 

 raria, but we often do not know the taxonomic 

 values of these characters. Bernard in order 

 to escape from the species dilemma, proposed 

 to give the various forms recognized by him 

 geographic designations: the collection from 

 each area is divided into as many forms as 

 possible; each one of these forms is given the 

 genus name, followed by the name of the geo- 

 graphic area from which it comes ; this in turn 

 is followed by two numbers ; the first, in paren- 

 theses, indicates the number of forms known 

 from the area; the second, which particular 

 form of the series is meant. To illustrate the 

 method, in Volume V. of the " Catalogue " six 

 kinds of Goniopora are recorded from north- 

 west Australia; these are designated Oonio- 

 pora Northwest Australia (6) 1, (6) 2, (6) 3, 

 (6) 4, (6) 5 and (6) 6. Each form is de- 

 scribed in detail, and when possible, figures 

 are given. No attempt is made to determine 

 the systematic value of the forms; that is left 

 to future work. This method can be com- 

 mended, as it furnishes a certain number 

 of detailed descriptions and figures and ac- 

 curate geographic data to subsequent students. 



and thus will aid them in unraveling problems 

 of coral systematics. 



In looking over Bernard's work, it appears 

 that he has not by the careful study of the 

 variation of different characters in the same 

 specimen, nor has he by carefully comparing 

 the variation of characters in different speci- 

 mens, attempted to discover the systematic 

 importance of the structural features de- 

 scribed by him. As the whole problem of 

 systematic zoology can be resolved into the 

 ascertaining of the relative values of the char- 

 acters possessed by the organisms under in- 

 vestigation, the reviewer feels that more in 

 the line of determining these values was to 

 have been expected of Bernard's work. 



In the opinion of the reviewer, Bernard in 

 his last volume carries his geographic idea 

 entirely too far. He says : 



It is in keeping with the conclusion at which 

 our work has brought us: that while free-living 

 organisms with highly developed powers of loco- 

 motion, such as fish or birds, may spread over 

 the surface of the globe, and thus be largely inde- 

 pendent of locality, this is not the case with ses- 

 sile forms as highly developed as the corals. 

 Whatever the " species " of corals may be, we know 

 nothing about them, and can know nothing about 

 them until we study them by means of experi- 

 mental cultivation. We only know the local forms. 

 Hence local forms are the only available units 

 with which we can do solid work. We have to 

 study them with a view to arranging them into 

 larger groups extending over larger areas. * * * 



It now seems probable that the forms of the 

 Indo-Pacific region will ultimately have to be 

 divided into smaller groups corresponding to defi- 

 nite areas: for instance the Red Sea forms will 

 be found to have characters peculiarly their owa. 

 A dim perception of this was pointed out by tha 

 author — ^but, as he now thinks, quite misunder- 

 stood — with regard to the genus Turiinaria (see 

 Vol. 11., 1896, p. 18, last paragraph). If this can 

 be established generally, a great reform in the 

 classification of corals can not be long delayed. 

 All purely imaginary groups such as species or 

 morphological " forms " of indefinite distribution 

 will be abandoned as units, and the corals will 

 have to be treated as we treat races of men— as 

 factors in the areas ichich they inhabit, and upon 

 the conditions of which they largely depend for 

 their peculiarities. 



