184 SELECTION. 



somewhat intermediate in size between Mus rattus and Mus 

 concolor, black in colour, with a very pronomiced brilliant green 

 lustre, and they had extraordinary long tails. 



Lloyd in his books on the "Growth of groups" gives num- 

 erous i nstances of such small, unif onn, aberrant populations 

 of rats, which can only be classed as decided species of a very 

 uncertain permanency. 



In all these cases, the species existence will come to an end, 

 when the special causes which afforded it the necessary isola- 

 tion from inter-crossing with the multitude of t37pical individ- 

 uals, will cease to exist. And this circmnstance shows us 

 that in the origin of species mere temporary isolation of a 

 small group with a high potential variability is not enough. 

 The combination of causes, the chance, which brought the 

 first colonists to the spot where they could found a new tent- 

 ative species, will eventually bring so many individuals of the 

 old species there, that the new group wiU cease to exist. And 

 it is clear that a species can only keep on existing if it can ex- 

 tend its range, so that adverse circumstances, which annihilate 

 it in certain regions, will leave sufficient munbers of individuals 

 to continue the species. A new species which differs from a mul- 

 titude of typical individuals cannot extend its range into the 

 territory of related forms, unless there are circumstances 

 which hinder free crossing. 



From the work of the systematicians with such animals as 

 smaU rodents, it foUows that such local species exist, which 

 cannot coexist in one locaUty, which differ from each other in 

 non-essential, non-adaptative characters, and which are 

 bound to a certain more or less circmnscribed area by the 

 existence of other local species of the same group in neighbour- 

 ing locaUties. In these circumstances it depends simply upon 

 the number of genes concerned in the differences envisaged, 

 whether we will find sharp or gradual demarkation between 

 the characters which distinguish these sub-species. 



On the other hand, it is possible, that a group of organisms 

 which is temporarily cut off from random crossing with the 



