412 Dr. F. A. Dixey on Mr. Merrifield’s Experiments 
among the features induced or revived by altered tem- 
perature-conditions, there is at least a residuum which 
must have owed its first origin to causes other than the 
direct action of temperature on the organism. Nor, 
again, are these to be considered as cases of “arrested 
development”; for the stages reproduced are stages in 
the phylogeny of the species, not in the ontogeny of the 
individual. 
If, then, these revived features are really ancestral, - 
how is their revival to be accounted for? The whole 
subject of reversion abounds with difficulty. An expla- 
nation commonly offered is that the characters last 
developed in the history of a species, or of an individual, | 
are less stable than those that have a longer history 
behind them, and that have become firmly established, 
under the operation of a long-continued process of 
heredity. Any disturbance—such as an exceptional con- 
dition of temperature—of the normal course of growth, 
may therefore be expected to act in the first place on the 
newer and less stable features, interfering with their 
usual line of development, and shaking back the species 
as it were to an earlier and more firmly-founded stage 
of its development—just as in an earthquake the freshly- 
built wing of a house, where the mortar was not yet dry, 
might fall and leave the older portions standing. Such 
an explanation, however, is in itself at best but partial, 
for it gives no real reason why the newer features should 
be less stable than the old; and indeed it comes to little 
more than restating the difficulty in another form. 
The two attempts to find a more definite explanation 
of reversion which may be said at present to hold the 
field, are those which pass respectively under the names 
of Darwin and Weismann. If the Darwinian assumption 
of centripetal “ gemmules” be granted, the commonest 
case of reversion, that namely which results from hybridi- 
zation, especially between recently-established species, is 
capable of explanation under the hypothesis of pan- 
genesis. But it may be questioned whether pangenesis 
as stated by Darwin is capable of accounting for such 
cases as the present, inasmuch as in them the condition of 
full maturity is almost reached before the introduction of 
the modifying disturbance. Although the ovum from which 
the individual has originated may under the Darwinian 
hypothesis have contained numerous gemmules of an 
