VARIATION IN SPACE AT PRESENT EPOCH 133 



tionally favourable conditions, which exceed in 

 some measure the ordinary term of their growth. In 

 these different cases the influence of the nature of 

 the environment on this variation becomes perfectly 

 evident, thanks to this method, which leads us 

 directly to the trinominal nomenclature,* that is to 

 say, to the necessity of using three names the 

 name of the genus, name of the species, and the 

 name of the mode of variation to designate a 

 shell. Take as an example Helix striata prcematura. 

 In view of the very great polymorphism of certain 

 species of land shells, does a criterion exist which 

 allows us to fix the point where variation stops and 

 where the next species begins ? In a word, is it pos- 

 sible to give a precise definition of a species ? This 

 is an old question, often discussed, and never de- 

 finitely solved. A purely morphological definition is 

 subject to error for the reason that polymorphism 

 sometimes leads, between the subjects of two good 

 neighbouring species, to an inversion, or, at least, to 

 an equalization, of one or several of the normal and 

 distinctive characteristics. Thus a big-bellied mode 

 of the Helix acuta may resemble, to the point of 

 being confused with it, a Helix ventricosa. To 

 guard against such errors, it is necessary to look 

 at the average characteristics of each colony, 

 rather than the individual ones of this or that 

 subject. But the naturalist has yet at his disposal 



* I am entirely of the opinion of Coutagne on this point, and 

 think that the adoption of a trinominal nomenclature is the sole 

 means of stemming the rising flood of so-called new species, described 

 without check and at the chance caprice of anyone, which threatens 

 to transform descriptive natural history into a veritable Tower of 

 Babel, where no one in future will be able to understand his neighbour. 



