66 'WHAT IS A SPFXIES^' 



This idea of a species is clearl)- expressed b)- Sir 

 William Thiselton-D)'er, when he speaks of the older 

 writers who employed ' the word species as a designation 

 for tlie totality of individuals differing from all others by 

 marks or characters which experience showed to be 

 reasonably constant and trustworthy, as is the practice 

 of modern naturalists'.' 



This conception of a species is founded upon transi- 

 tion. Whenever a set of individuals can be arranged, 

 according to the characters fixed upon by the s)stcmatist, 

 in a series without marked breaks, that set is regarded as 

 a species. The two ends of the series may differ im- 

 mensely, ma)- diverge far more widely than the series 

 itself does from other series; but the gradual transition 

 proclaims it a single species. If transitions were all 

 equally perfect, of course there would be no difficulty. 

 But transitions are infinite in their variet)- ; while the 

 subjective element is obviously dominant in the selection 

 of gaps just wide enough to constitute interspecific breaks, 

 just narrow enough to fuse the species separated by some 

 other writer, — dominant also in the choice of the specific 

 characters themselves.- Looking back upon the interval 

 between Linnaeus and Darwin, it seems remarkable that 

 the mutability of species was not forced upon systematists 

 as the result of their own labours. It is astonishini^ that 

 many a naturalist was not driven by his descriptive work 

 to the conclusion which Darwin stated to Asa Gray on 

 July 20, 1S56 : ' — as an honest man, I must tell you 

 that I have come to the heterodox conclusion, that there 

 are no such things as indepcndentl)' created species — 

 that species are only strongl)- dehned varieties.' •' 



For, as I have said above, every describer of species 

 made continuit)- and transition in characters the test of 

 a variety, discontinuity the test of a separate species. 



' loc. cit. p. 370. 



^ How important this choice may he is well shown by Karl Jordan in 

 Novitati's Zoologicae, vol. iii, Dec. 1896, pp. 428-30. Characters are 

 subject to independent variation as well as correlated variation. Ilence 

 there will often be the widest discrepancy between the transitions con- 

 structed by naturalists making use of different characters. 



' Lift and Letters^ vol. ii, p. 79. 



