BUIXETIN 5!). T7NITED STATES NATIONAL Ml'SEUM. 5 



Tliese 9 forms arc not niiuu'd, and no attempt is made to identify them with forms 

 from another area. Each one is tfiven a number, and they ai'c designated Pi>r'iteH 

 Sandwich Islands i, |, |, \, etc., the denominator indicatiny- tiiat lie has recoi^nized 9 

 forms from the Sandwich Islands, the numerator indicatini>' which one of the 9 is 

 meant. Should additional forms be discovered later, the numerator and the denomi- 

 nator could be increased. In this way Mr. Bernard hoped to recxjrd all the forms 

 known from any particular area, and avoid expressing' an opinion as to their group- 

 ing into species. 



This proposition is so iconoclastic, in a \\ay, that it was only to be expected that 

 it would arouse considerable opposition. The criticism published by Mr. rl. Stanley 

 Gardiner is the one most deserving of attention." Bernard's proposition is not 

 absurd, for without large series of specimens for comparison it is impossible to work 

 out the limits of variation; as he did not have such series, he recognized that if he 

 were to describe as a species every specimen or group of specimens not coimected 

 witli others by intermediates, he might be, or even probal)ly would be, describing, 

 as species, specimens that formed parts of a continuous s(>ries. Gardiner later real- 

 ized more fully the dithculty with which Bernard was contending, and was not onl}^ 

 more lenient in his criticism, but acknowledged * that Bernard had probably handled 

 his material in the best possibl<> manner. (Jardiner, in his own work, however, 

 thought that he could recognize specitic limits with sutiicient clearness to continue 

 the use of the binomial system. 



Whatever oi)inion may be held as to Bernard's metliod of designating the forn)s 

 tiiat lie records, it is undeniable that he has made valuable contributions to our 

 knowledge of the Madreporaria. 



My own position regarding nomenclatorial problems is somewhat ditl'ercnt from 

 that of Mr. Bernard. There are comparativ(dy few of the inductions of science that 

 can properly be regarded as more than tentative. In the matter of species, we are 

 not yet decided as to what a species is. The conclusions of de Vries, should thev be 

 correct, would fuinish a working l)asis, but the mutation theoty is at present onlv a 

 valuable working hypothesis. All of our descriptive systematic work, therefore, must 

 1)0 regarded as only tentative. However, as it has tlic double object of recordino- the 

 manifold forms assumed l)V organic nature and attemjitii.g the grouping of those 

 forms according to the degree of their kiiislii|). w(> are jiistitied in naming forms as 

 species: for, although we may not know witii positiveness their actual svstematic 

 value, the determination of the relationship is attempted. The detinition of species 

 is only an attempt to divide organisms into groiqis in accordance with the detinition 

 of the term "•sjiecies"' previously given. I am in favor of atta<hing names to tiu>sc 

 tentative groups, as names are more serviceal)l(>, in discussion than symbols, especially 

 when the symbols are complicated. 



Great pei-plexity has been experienced in preparing the succeeding pa[)er, as to 

 how the results ol)tained by the study of the material in hand should be presented. 

 The term "species" is used in accordance with the detinition given. The word 



«0n the Unit of Cla.»isific.ation for Systematic Biology. A reply to Mr. Bernard. Procee<iings, 

 Cambridge Philosophical Society, XI, 1902, pp. 423-427. 



^ Madreporaria, Fauna and Geograpliy of the Maldive and l.aciadivc .\rchipelagoes, II, pp. 75t»-7.57. 



