ISOLATION 181 



part of the neck are white. Finally the young of 

 P. colensoi have the neck dark as in P. campbelli. 

 So that we arrive at the conclusion that P. camp- 

 belli is the progenitor of P. colensoi. 



Here we see one species changing its white abdo- 

 men for a black one ; another changing a black 

 abdomen into white. One species acquires a white 

 bar on its neck, another loses it. P. carunculatus 

 has lost the crest on its head which its ancestors 

 certainly w r ore ; while several have changed the 

 colours of the skin between the eyes and the mouth. 

 These changes cannot be due to the action of ex- 

 ternal conditions, and we cannot suppose that they 

 are due to utility, for the conditions under which the 

 birds live are remarkably similar. 



In the Auckland Islands we see the evolution of a 

 new species going on. The parent form, P. camp- 

 belli, is rare, and confined, I believe, to the south 

 part of the island ; while P. colensoi is extremely 

 abundant. There seems to be no need for recogni- 

 tion marks either with P. colensoi or P. breviro- 

 stris. In both cases the new variation is simply 

 swamping the older form. 



Some of the specific characters, such asi the crests, 

 the bright colours of the skin on the face, and pos- 

 sibly the colouring of the feet, are decorative, and 

 probably due to preferential selection ; while other 

 marks, or loss of marks, are due to reversion. How- 

 ever, the idea of preferential selection does not alto- 

 gether satisfy me ; for why should a preference for 

 the same character be shewn by successive genera- 

 tions? There seems to be nothing to make 



