INTRODUCTION 5 
it has no vertebrae; on the other hand, it has a heart of an am- 
phibian rather than of a fish type. 
Some of these considerations may be represented graphically 
as follows: let us take a species D that has an ontogeny A, B, C, D, 
and suppose that this species evolves successively into species 
EK, F, G, H, ete. When evolution has progressed a step, to E, 
the characters of the species established develop directly from 
the ovum, and are therefore, in some way, involved in the com- 
position of the latter. All of the stages of the ontogeny leading 
up to E are modified, and we can indicate this in the ontogeny 
LAB CD of FE. as in line 2; similarly, when evolu- 
2a Ae BSCE DME tion has progressed to species I, seeing 
Bie Vase Bil Os Dies Deal th that the characters of F now develop 
4A" BC? D? b? EUG directly from the ovum, all the onto- 
5. A* Bt Ct D* E? F? GH genetic stages leading up to F are modi- 
fied, line 3. And so on for each successive advance in evolution, 
lines 4 and 5. It will also be noticed that the terminal stage D of 
species 1, becomes a successively earlier ontogenetic stage of species 
2, 3, 4, 5, etc., and moreover it does not recur in its pure form, 
but in the form D' in species 2, D? in species 3, etc. Nowif the 
last five stages of the ontogeny of species 5 be examined, viz., 
D*, KE, F?, G', H, it will be seen that they repeat the phylogeny 
of the adult stages D, E, F, G, H, but in a modified form. 
This is in fact what the diagram shows; but it is an essential 
defect of the diagram that it is incapable of showing the character 
of the modifications of the ancestral conditions. Not only is each 
stage of the ancestral ontogenies modified with each phylogenetic 
advance, but the elements of organization of the ancestral stages 
are also dispersed so that no ancestral stage hangs together as a 
unit. The embryonic stages show as much proportional modi- 
fication in the course of evolution as the adult, but this is not 
so obvious owing to the simpler and more generalized character 
of the embryonic stages. 
The recapitulation theory as outlined above is obviously a 
corollary of the theory of organic descent; it was in fact developed 
in essentially its present form, soon after the publication of the 
“Origin of Species,” by Fritz Muller and Ernst Haeckel. But 
the data on which it was based were known to the earlier embry- 
ologists; and Meckel, for instance, insisted very strongly on the 
resemblance between the ontogenetic and the taxonomic series 
