variety, every separate degree of colour which could 

 possibly be found between the outer limits, seeing that 

 (increasing, as they did, in an even ratio) no one could 

 be tabulated in preference to another. 



This however is an example in which the rate of altera- 

 tion (so far as colour is concerned) is equal ; and one 

 therefore in which the extreme end of the series can be 

 alone singled out as the aberration to be specially noticed. 

 It sometimes occurs that, between the two extremes, 

 there are several nuclei, or centres of radiation, to which 

 the name of varieties may be legitimately applied, in- 

 asmuch as they may possess a series of characters which 

 do not, all, in combination, progress evenly ; and which 

 consequently stand out as it were, to a certain extent 

 isolated, from the remainder. 



As a corollary arising out of these remarks, it would 

 seem to follow that even small differences should be re- 

 garded as specific ones so long as the intermediate links 

 have not been detected which may enable us to refer 

 them to their nearest types. In a general sense, I 

 believe that it would be proper to do so : nevertheless 

 there are instances, the results, for example, of isolation, 

 in which abrupt modifications may be a priori looked 

 for ; and in which our judgment must be regulated by 

 our knowledge of the local circumstances which may be 

 reasonably presumed to have had some influence in pro- 

 ducing them. The consideration of these, however, and 

 other kindred questions, must be deferred to a subse- 

 quent chapter of this work. 



