A CRITICAL KE-STATEMENT OF THE BIOGENETIC LAW. 87 
only are capable of begetting, the thing begotten is built up of successive 
adults. That " cenogenetic '' interpolations*, without any adult ancestral 
significance, are a normal feature of almost every life-cycle, is often verbally 
admitted, though the recapitulationist rarely realises how profoundly such 
interpolations may affect his phylogenetic conclusions, and how dangerously 
subtle become his arguments when engaged in proving that such larvse as the 
Trochosphere, the Nauplius, and other distributive stages of the life-cycle 
are at bottom " recapitulative.''' I propose to deal more fully with the origin 
and significance of such larval forms in another communication ; but, to 
illustrate the principle, I may add one example of the origin of an inter- 
polation in the most progressive of sequences known to us, viz. that of 
Vertebrates. There is no doubt that Birds are descended from Reptiles. 
It is beyond question that Reptiles are hatched in a form and with a somatic 
organisation which is that of a miniature adult Reptile in all respects. Yet 
the Bird is hatched in a form and with sundry details of organisation 
different from those of the adult, e.g. its downy plumage. Now, the 
"typical" down-feather is an open hollow tube, splayed out at its free 
extremity into a ring of soft barbs (or barbules) of equal size, and I ask if 
such a tubular feather is to be regarded as an intermediate stage in the 
phyletic derivation of feathers from scales. I submit l;hat there is not a scrap 
of evidence, or of probability, that any adult ancestor of Birds, along the 
whole route from Reptiles to Sparrows, was ever clothed in anything except 
scales, feathery scales, and finally contour-feathers. The chick is an inter- 
polation in the life-cycle of Birds, and its down is a "secondary" modifi- 
cation of complete contour-feathers. The Duck, the Fowl, and the Pigeon 
represent three successive grades of differentiation in the phylogeny of 
Birds. Anyone who will examine under a microscope the nestling-down of 
these three birds in the order mentioned, will see that they exhibit successive 
phases in the degradation (a) of the primitive rachis of a contour-feather 
and (h) of the barbs of such a feather, i. e. that the chick stage, with its 
peculiarities of organisation, has been evolved, step by step, within the group 
of Birds alone, and is an interpolation that has no relation with, and throws 
no light on, the prse-Avian adult ancestry, or on the way in which scales 
wei"e transformed to feathers f. 
* E.g. Weismann's discussion of the evolution of markings in larvas of SphingidaB, much 
of whicli is probably sound, though unnecessarily complicated by the assumption that 
primitive longitudinal markings have been " shunted back " into earlier stages of the 
ontogeny, instead of being simply replaced in the later stages by patterns more suited to 
increased size or special conditions of exposure (1904, pp. 177-185). 
I The subsequent publication of Prof. Cossar Ewart's valuable paper on " The Nestling 
Feathers of the Mallard " (P. Zool. Soo. 1921) renders this discussion inadequate arid I hope 
to amplify it. In the meantime I would merely remark that, on the relation of feathers to 
scales, the association of several feather-germs with single scales on the foot of the Owl is 
no disproof of my thesis, since the feathers here represent a secondarj' extension, like that 
of the scales on the head of Ceratodus. 
