THE TRANSMUTATION OF SPECIES. 245 



change from the very earliest historical times onwards. 

 Though, for readily intelligible reasons, seen in a more 

 marked form in domestic animals than in wild ones, such 

 groups of differing individuals are seen in a vast number 

 of species of animals from the lowest to the highest. 

 Most ' species ' of animals, therefore, include one or more 

 'varieties;' .and this is a phenomenon quite as well seen 

 in the vegetable kingdom as amongst animals. 



Now, the question arises at what point do the differ- 

 ences which distinguish a 'variety' from a 'species' 

 become so pronounced, that we conclude that we have a 

 fresh species to deal with, and not a mere variety of an 

 old species ? This is a point upon which naturalists have 

 not as yet succeeded in laying down any fixed rules. 

 From the nature of the problem, it is extraordinarily 

 difficult to detect any underlying principle to guide us in 

 practice in deciding between species and varieties. 

 Hence, in what are called 'variable' genera (such, for 

 example, as the genus Rubus, including the brambles, or 

 the genus Salix, containing the willows), observers have 

 never been able to agree as to the precise number of 

 species which exist, even in a small country like Britain 

 since what one observer sets down as two species, 

 another equally competent authority will regard as being 

 only a species and its variety. The test usually adopted 

 by naturalists in distinguishing between varieties and 

 species is what has been called the 'physiological test.' 

 That is to say, when the individuals of a given assem- 

 blage of animals or plants are fertile, and are capable of 

 giving rise to fertile offspring, then they are usually 

 regarded as constituting a single species^ however greatly 



