CONTINUOUS VERSUS DISCONTINUOUS VARIATION 



groups in nature, and that is generally the case, the conventional classi- 

 fications may be said to be natural. But, to the extent that the rank 

 accorded the various groups is arbitrary, the system of classification itself 

 is arbitrary rather than natural. Taxonomists generally are agreed that, 

 while the discontinuities of the higher groups are real, their assignment 

 to systematic categories is primarily a matter of convenience. Linnaeus 

 himself said this of the higher categories, but the species he regarded as 

 a real thing, each one a specially created unit. The replacement of the 

 archetypal concept with the evolutionary concept has not been accom- 

 panied by an abandonment of Linnaeus' idea of the definiteness of species. 

 As Bateson has said, "Though we cannot strictly define species, they yet 

 have properties which varieties have not, and . . . the distinction is not 

 merely a matter of degree." 



It becomes obligatory then to discuss the nature of this unique taxo- 

 nomic unit, the species. In Chapter 3, species were defined as kinds of 

 plant or animal the individuals of which differ "from each other only in 

 minor traits, except sex; sharply separated in some traits from all other 

 species; and mutually fertile, but at least partially sterile when crossed to 

 other species." 



Magnitude of DifFerence. Many other definitions of species have been 

 published, and all of them, including the above, are unsatisfactory in the 

 sense that they do not provide a basis upon which a practical taxonomist 

 can decide whether two similar groups are distinct species or only sub- 

 species. Some have tried to specify the degree of difference which is neces- 

 sary to separate good species, but this is not practical not only because of 

 the difficulty of formulating a quantitative expression, but because the 

 degree of difference between species in some groups seems to be very 

 much greater than in other groups; and because some undoubtedly good 

 species, as Drosophila pseiidoobscura and D. pershnilis, show very little 

 morphological difference, while different races of other undoubtedly 

 single species, like man, show very pronounced difi^erences. It appears that 

 the degree of difference is much less important than the constancy of 

 difference, that is, the discontinuity between the groups. 



Discontinuity and Interspecific Sterility. But even strictly discontinu- 

 ous differences need not indicate specific boundaries. Differences condi- 

 tioned by a single gene will show complete discontinuity if dominance is 

 complete ( for either allele ) , and these differences may be of considerable 

 magnitude. For example, the mutant tetraptera in Drosophila is character- 

 ized by the development of two pairs of wings, much as in the dragonflies 

 and many other orders of insects. The type of discontinuity of greatest 

 interest is that which results from the inability of related species to inter- 

 breed, the phenomena of interspecific sterility and hybrid sterility, con- 

 cerning which there is an immense literature. Many definitions of species 

 have emphasized this. It has the merit of being very generally true. But 

 it also has some faults. At the outset, taxonomists must of necessity do 

 most of their work with preserved specimens, and hence a sterility barrier 

 would be difficult to use even if the validity of the concept were com- 

 pletely agreed upon. But there are some facts which cast doubt upon the 



265 



