404 NATURAL SCIENCE [December 
gressed backwards towards a more ancient order of things; and 
at every stage of the ontogeny structures are absent, which were 
present in the phylogeny because they were then useful, but which 
since underwent complete retrogression, because they subsequently 
became useless. Here, then, we have the explanation of the 
fact that ontogeny is only a very vague recapitulation of 
phylogeny. Doubtless if a higher animal, a man for instance, 
lived during his ontogeny in a succession of environments similar 
to those in which his race was evolved, his ontogeny would much 
more exactly recapitulate the phylogeny than it actually does, 
for in that case structures, which had been useful during the 
phylogeny, would continue to be so during the ontogeny, and so 
would be preserved. But consider how vastly different is the 
environment, in which the embryo of man develops, from the en- 
vironments in which his race evolved. The embryo develops in the 
uterus, but its free prototypes struggled each for itself in a world 
full of enemies, full of eliminating agencies. How many parts, 
therefore, have become useless to the embryo, which were useful to 
the prototypes! How vast is the field in which retrogression has 
worked! Is it any wonder, then, that the ontogeny of man is only 
a vague recapitulation of his phylogeny ? 
Reversion, then, is the necessary complement of evolution, and 
without it there could be no evolution, except of the simplest kind. 
Without reversion there could be no planing away of the number- 
less useless variations which occur during, and especially at the end 
of the ontogeny, nor of all those structures, which, though useful 
during some part of the phylogeny, became useless later. Without 
reversion, therefore, a species would soon become so burdened with 
useless variations and structures as to be incapable of existence. 
Reversed Selection could not cause the elimination of all these 
useless and burdensome characters; for no matter how burdensome, 
and, therefore, worse than useless, they are in the aggregate, separ- 
ately they are so little burdensome that Reversed Selection could 
not act. It could not act on them in the aggregate, for this would 
mean that in some individuals they would be present en masse, 
whereas they would be absent en masse in others; and this, of course, 
we know is not the case. Moreover Reversed Selection causes a 
retracement, not a lapsing of characters. It therefore works at a 
double disadvantage as compared with ordinary Natural Selection, 
and, as a consequence, can effect comparatively little. No exten- 
sive examples of such retracement are in fact known to us in 
Nature. Again, without retrogression, the recapitulation of the 
phylogeny in the ontogeny would be impossible, and, for this reason 
once again, evolution would be impossible. _ For, were there no 
retrogression, the prototypes of the phylogeny would necessarily be 
