2()(j OKKilX OF FEATIIKKS FltOM SCALES. 



feathers largely by a complicated series of incisions of an 

 upgrowth from the scale. Further, it is possible to show that 

 the scale u{)growth which becomes changed into a feather is really 

 comparable with the overlap found in the imbricating scales of 

 many lizards and snakes; also that in its formation the feather 

 has a dorso-ventrality comparable with that of the overlap of 

 scales, the outer, dorsal aspect of the scale representing the shaft 

 of the feather, and the under ventral surface the after-shaft of 

 the feather. This homology established, it is also easy to show 

 that the overlap of scales is comparable with claws and nails, 

 and that the unguis of the latter represents the shaft of feathers, 

 and the sub-unguis the after-shaft. Longitudinal invasions of 

 the dermis into the epidermis take place in all of them, though 

 they are sufficiently deep to result in the splitting or fraying of 

 the horny material only in the case of feather formation. 



The second view is that which we seem compelled to accept 

 on a strict adherence to the factorial hypothesis, upon which 

 modem Mendelism is based. This assumes that every character 

 in the body has discrete representation in the germ plasm; that 

 new characters appear as a result of changes in the germ plasm; 

 that they are mutations or saltations, not a gradual transition 

 from something previously existing to something new. In so 

 far, then, as the relationship of the feather to the scale in the 

 ostrich really represents what has occurred in the evolution of 

 feathers, we are to regard the feather as a wholly new structure 

 which has appeared in association with a scale, but that in its 

 origin and nature it is something apart from and independent of 

 it. It is not that ttie scale has been transformed into a feather. 

 There is no such thing as homology between the scale and the 

 feather, as the term is usually employed. Each has separate and 

 independent representation in the germ plasm. Thus, instead 

 of regarding feathers as homologous with the overlap of the scales 

 of reptiles, and also with the claws of reptiles and birds and 

 the nails of mammals, they are to be considered as distinct 

 mutations which have appeared in the course of evolution of birds 

 from reptiles, and their association with scales and resemblances 

 to other horny structures are altogether incidental. Coincident 

 with the appearance of feathers we have a disappearance of 

 scales, either by loss of factors or their inhibition, apart from 

 the relics on the legs and toes where the two continue together. 

 It is manifest that the application of the conception of the 

 discreteness of characters in this fashion raises the whole ques- 

 tion as to the meaning of the comparisons instituted in studies 

 of homology and comparative morphology. According to the 

 factorial view, a modification of an existing structure by the 

 addition of a new feature is not a transformation of the old, 

 but represents something wholly new. Homology, however, is 

 based upon the idea of transformation; a structure liaving certain 

 characters has been changed into another with other characters, 

 but without losing its identity. We may take, for example, 

 the visceral arches of vertebrates. In comparative anatomy we 

 attempt to homologise the cartilaginous arches in such a fonn 



