VARIATIOX IX RIDS OF PEXGUIN. 288 



the inost part, the anatomical and palaeontological evidences of 

 evolution seem to call for successive series of germinal changes 

 in the same direction, but it may well be doubted whether this 

 is their real interpretation. Apart from the influence of selection, 

 no experimental evidence is forthcoming in favour of germinal 

 changes taking place in an3' successional order. Mutations, as 

 we know theni, are haphazard and independent of one another, 

 following in no definite order; each appears to be discrete and 

 apart from the others. Often, however, the germ factors con- 

 trolling certain characters seem to be in a more mutative state 

 than others, and the germinal changes and resulting mutations 

 are not necessarily the same or of the same degree in different 

 individuals. Thus, we may assume that the germinal factors 

 controlling the tenth rib of the penguin are, or have been, in a 

 more changeable state than those controlling the remainder of 

 the ribs, and this, along with the interbreeding of the different 

 individuals, has given us the great variation now encountered. 

 The variation is of such a character that it suffices where a 

 sufficient number of individuals are available to produce a con- 

 tinuous series extending from one extreme to the other. As a result 

 of interbreeding, these variations do not now necessarily represent 

 the original germinal changes: they are intermixtures of the 

 (liiTerent factorial states ; much less do they indicate that the 

 changes wei'i' in an ordinal so(pu'nce, leading to the loss of the 

 rib in a continuous fashion. 



It is manifest that the last ])air of ribs coming, as it were, 

 at the boundary between the thoracic and abdominal regions of 

 the body may be subject to different stimuli from the other 

 members of the costal series, and would respond more to any 

 changed' habits of the bird. And if we could follow Mr. J. T. 

 Cunningham,"" and hold that these stimuli produce hormonic 

 secretions capable of influencing the germ plasm, then we have 

 a means by which the variations may have been induced and 

 limited to the structure in question. 



If no selection of any kind is introduced, and no further 

 mutations occvu', penguins will presumably continue to give all 

 the variations in the tenth rib here described, without the latter 

 ever being suppressed. If selective breeding could, however, be 

 instituted it would be possible on the one hand to produce a race 

 in which the tenth rib would be altogether suppressed, and, on 

 the other, a race in which it would be completely developed along 

 with a portion of the eleventh. These would be races in which 

 germinal purity had been attained by the elimination of other 

 divergent factors, a condition very different from that of the 

 penguin at the present time. 



" Hormones and Heredity," London, 1921. 



