ORIGIN OF FEATHERS FROM SCALES. 267 



as the dogfish with the bony arches of the teleost, and it is then 

 sought to compare these with diverse bony structures in the 

 amphibian, bird and mammal, the details of which are familiar to 

 every student of the vertebrates. We know that the cartilage 

 has become replaced, and assume that it is represented by the 

 bones of the teleost, and that these in their turn have become 

 transformed into bones of different form in the amphibian, 

 reptile, bird and mammal without loss of identity. Thus, the 

 pterygoid is supposed to represent the same bone throughout the 

 series, whatever be its form, and also to be represented by a 

 part of the upper jaw of the cartilaginous fish. "With some show 

 of justification we can point to the fact that at their earliest 

 stage of development, the cartilaginous stage, the arches of all 

 vertebrates are practically alike, and then assume that the bones 

 derived from these and occupying the same relationships 

 correspond throughout the vertebrate series. 



On the factorial hypothesis, however, we can just as well 

 regard the bones of the teleost arches as entirely new structures 

 which replace the earlier cartilage, represented, as they must 

 be, by new factors in the germ plasm. It is true they appear 

 to be a transformation of the latter, but this has no further 

 meaning than, on the same hypothesis, has the replacement of 

 scales by feathers in the ostrich. They are not homologous 

 structures, though they occur in the same position and have 

 essentially the same relationships, if they can be held to be the 

 expression of germinal factorial differences. Similarly with the 

 modifications of the visceral arches in the higher classes of the 

 vertebrates. The bones which replace the early cartilage are not 

 necessarily the same as those replacing it in the teleost; unless 

 they represent modifications due to different somatic influences 

 they are new formations with their own germinal representation 

 distinct from that in teleosts. If each change in a structural part 

 indicates a germinal change, and takes place apart from what 

 has gone before, the term homology as usually understood has 

 no application; each modification represents a new character and 

 is only incidentally associated with the old. 



We can only accept gradual morphological transformation as 

 a basis of evolution if we agree with WaiTen* that factors them- 

 selves undergo change, as he concludes from his expei-iments 

 on the crossing of foxgloves, or accept the hormonic hypothesis 

 of Cunningliam, i which attempts to show how a character may 

 arise or be modified by the production of internal secretions. 

 Both these afford a suggestion as to how the old may be gradually 

 transformed into the new, and thereby give us a basis for 

 comparative morphology. 



If, then, we follow strict Mendelian principles, we must 

 conclude that feathers in birds have originated de novo from 

 independent factorial germinal changes, and that their association 

 with scales is only incidental; whereas, if we apply the recognised 

 principles of homology, we must hold that feathers have evolved 



* South African Journal of Science, vol. xviii, 1922. 

 t " Hormones and Heredity," London. 1921. 



