V. GRANT 65 



Davidson (1954) has pointed out, "If we wish to argue that 

 species have an objective reality, we may select examples such 

 as Ginkgo. If we wish to argue that species do not exist in nature 

 at all, we may select such examples as the willows." The species 

 of willow are notoriously confused by natural hybridization, 

 whereas this factor can scarcely affect Ginkgo biloba, which is 

 the sole living member of its phylad. 



The frequent occurrence of the willow pattern of evolution, 

 as contrasted with the ginkgo pattern, is perhaps reflected in the 

 species concept held by so many of the nineteenth century plant 

 hybridizers. Their practical experience impressed upon many of 

 these early students, Naudin, Nageli, Darwin, Mendel, Kerner, 

 De Vries, and others, as it has also on many modern botanists, 

 an explicit disbelief in the notion that the species as a unit is 

 qualitatively distinct from the race. 



The failure of conformity of the evolutionary developments in 

 certain groups of plants to the biological species concept in its 

 present formulation suggests the need for a reexamination of our 

 definitions. It sometimes proves necessary to split a biological 

 unit into two or more units. As previously mentioned, the incon- 

 sistencies of the "concept of the individual" when applied to 

 clones forced a recognition of both individuals and clones; and a 

 splitting of the gene into functional and recombinational genes 

 is apparently in the offing. Similarly, the evidence derived from 

 evolutionary studies of higher plants indicates the existence of 

 a unit of interbreeding higher than the species. That unit is the 

 sum total of the species linked together by frequent or occasional 

 hybridization. 



Categories supplementary to the species have been proposed 

 by several authors. The coenospecies of Turesson (1922) and the 

 commiscuum and comparium of Danser (1929), being based on 

 the fertility relationships of species in the experimental garden, 

 are artificial categories of use only for the discussion of a special 

 aspect of natural populations. The coenogamodeme and syngamo- 

 deme of Gilmour and Heslop-Harrison (1954) are synonymous 

 with coenospecies and comparium respectively. The term micton 

 of Camp and Gilly (1943), although corresponding to a natural 



