A CEITICAL RE-STATBMENT OF THE BIOGENETIC LAW. 85 
the life-cj'cle alone effected. The Coelonterate had a larva fitted both to 
distribute it and to build it up. It must be changed so as to perform this 
double function for a quite different creature, probably of verj- different size, 
habits, and requirements. And then the changes at the larval end must be 
fitted and co-ordinated with the changes at the adult end, so that every phase 
of the life-cycle is modified in some way or other. Yet it is only this much- 
pruned Ccelenterate sequence that survives as building material out of which 
a specifically Annelid ontogeny may give rise to a Crustacean, and so on. 
Inevitably the Ccelenterate sequence in the Crustacean's ontogeny is reduced 
to the simplest terms, and is as far from " mirroring " any functional Ccelen- 
terate type, or the original mode of its formation, as possible. Nevertheless, 
the grades persist as stepping-stones from zygote to adult ; and, having been 
successively pruned of unessentials as they ceased in turn to furnish directly 
the equipment of the adult stage, they have become very constant features 
of the ontogeny in a long line of evolutionary progress. For there is an 
irreducible minimum beyond which even ontogeny cannot abbreviate. The 
zygote is always unicellular, the larva multicellular and fitted for swimming, 
and the adult a raultilaminate complex of interdependent parts ; and even 
the Ctenophore, with its elaborate prje-organisation of the zygote, cannot 
escape the rule that 8 = 4x2. Ontogeny repeats the necessary successive 
grades of ancestral differentiation, but no ontogenetic stnge is ever more 
than an immature adumbration of a particular adult type in the phyletic 
chain *. It reproduces those successive grades, not because successive adult 
types have been included in it, but because each ontogeny is a modification, 
within limits, of its predecessor ; and by those predecessors the phyletic 
chain of adults was organised and equipped. 
7. Thus Ccelenterate, Goelomate, Protochordate, Gnathostome, and Tetrapod 
are successive grades of differentiation both in the ontogeny and phylogeny 
of a Frog ; but at none of these grades does the ontogeny recall the fornl 
and structure of a possible adult ancestor. This is obvious enough in each 
of the first three grades ; and in the fourth, which is held to " recapitulate " 
the Fish, the tadpole lacks dermal skeleton (both scales and fin-rays), paired 
fins, and biting jaws, which the adult ancestral Fish undoubtedly possessed. 
The tadpole, in fact, is not a modified reproduction of an adult Fish-ancestor, 
but a modification of the larva wliich that ancestral fish undoubtedly 
possessed — still recognisable, in less modified form, in the larvae of Polypterus 
and Dipnoi to-day. In other words, the life-cycle of the Frog is a modifi- 
cation of the life-cycle of an ancestral fi-eshwater Fish ; and adjacent terms 
in the old life-cycle (larva and adult) have undergone parallel and correlated 
modifications, as well as some independent specialisations. 
» Cy. Von Baer (1828, p. 230): "Der Embryo geht nie durch eine andere Tierform 
tiindurch, sonderu nur durch den Indifferentzzustand zwischen seiner Form und einer 
anderen." 
