174 THE president's address. 



fishes, and bats with birds, and I need not waste your time by 

 pointing out how unnatural such a method of grouping would 

 be. Many perfectly logical classifications are possible, but only 

 one that is perfectly natural, and that one, very often, is not 

 perfectly logical in form. 



The practice of classifying living things into systems more 



or less logical may be as old as Adam, but it arose in a time 



when living species were supposed to be as fixed and immutable 



as those of the inorganic world. Tot sunt sj^ecies quot ah initio 



creavit infinitum ens was the dictum of Linnaeus, the father of 



modern classification of living things ; but even in his time the 



existence of natural and artificial classifications was recognised, 



though the meaning of the difference was first made clear when 



the origin of species, through descent with modification from 



other forms, became an accepted notion. It is now abundantly 



clear that natural groups can seldom, if ever, be defined by 



precise and rigorous verbal definitions. All that can be done 



is to construct for each group a more or less ideal and imaginary 



type of organism, possessing certain characters, none of which 



must be regarded as fixed or invariable ; what Delage and 



Herouard call a morphological type in their great work, La 



.Zoologie Concrete. If we must have verbal definitions of groups, 



then logic requires the insertion of the word "typically" before 



each character ascribed to them ; thus Tetraxonida are Sponges 



which possess, typically, four-rayed siliceous spicules ; Reptiles 



are, typically, pentadactyle Vertebrates. Only in one of these 



two ways can we effect a just compromise between the claims 



of strict logic and the exuberant versatility of Nature. 



To return now, after this somewhat long digression, to the 

 Protista. Can we recognise in them any prominent differences 

 of character which may serve for the purpose of subdividing this 

 vast group? As already stated, we abandon the time-honoured 

 distinction of animal and vegetable as inadequate and imprac- 

 ticable in these lowly organisms. What can we use instead ? 



I believe myself that there are two well-marked types re- 

 cognisable in the Protista, the one more primitive and older in 

 evolution, the other higher and leading on to the ordinary plants 

 and animals. As I have been at some pains to point out, the 

 existence of two such types does not preclude on the contrary 

 it postulates the existence in the present or past of every 



