172 Determinate Variation and Selection 



selection,' 'sexual selection/ etc., it is both unnecessary 

 and unwise to attempt now to call them all 'isolation.' 

 For if everything is isolation then we have to call each 

 case by its special name, just the same, to distinguish it 

 from others. 



There remains the question as to whether isolation, in 

 the broad sense of the restriction of pairing to members 

 of the same group, can result in specific differences with- 

 out any help from * selection ' of any kind. If that should 

 be proved,^ then there would be, it would seem, justifica- 

 tion for the term 'isolation' in evolution theory, with a 

 meaning not already preempted. This Professor Hutton 

 claims, with Romanes and Gulick. 



1 At present it is far from being proved. Cf. Professor Cockerell's review 

 of Romanes in Science, April 29, 1898. 



