THE EVOLUTION OF LIVING BEINGS. 23 



be uniformly black, while the young ones of the other 

 female will be uniformly grey. Breeding true to type is 

 consequently by itself no reliable test for specific purity 

 and consequently there is no certainty that a Jorda- 

 nian species is really a species; forms from which we 

 know nothing else than that they breed true to type, 

 may consequently not be designated as species but 

 must receive another name; as such I propose Jor- 



DANONS. 



The question remains : what is a species ? 



The great advance in hybrid analysis in recent years 

 which allowed us to find out the constitutional diffe- 

 rences existing between externally indistinctible forms 

 as the white mice, mentioned above, allows us to give 

 a definition of the species, viz : 



A species consists of the total of individuals of identical 

 constitution unable to form more than one kind of game- 

 tes. 



All specifically pure individuals are consequently 

 monogametic. The question remains how to test for 

 specific purity and the best applicable, although not 

 always reliable, test is probably the one suggested 

 recently by Davis: 



If two individuals crossed together give a uniform F u 

 these two individuals are specifically pure. 



The explanation is quite simple : each of them pro- 

 duces but one kind of gametes ; all children therefore 

 must have the same constitution and consequently 

 must be alike under like conditions. 



Some possible objections however must be conside- 

 red: 



