896 The American Naturalist. [October, 
by crustaceans, a physical weakening results, and new tissue 
must be built.up before the new shell can appear. A similar 
weakening is apt to appear in man during the development of 
puberty, and various other instances might be given. 
All this leads back to the question of atavism. The changes 
indicated may not be solely due to nutrition and stimulation, 
but may be controlled in a measure by the original germinal 
conditions, the degree of developmental vigor which exists in 
each of the molecular groups of the germ cell. If any of these 
is weakly constituted, or imperfectly organized, its general de- 
velopment may cease before the ultimate phase is reached, or it 
may be imperfect, and the resulting animal lack some part, asin 
the absence of a hand or arm. This may be the ordinary 
cause of the phenomena of atavism, the original weakness of 
the germ causing a cessation of development before the final 
stage is reached. This check seems often to occur at the level 
of some immediate ancestor, but occasionally acts at a consid- 
erably more remote stage. Again, weakness in a special region 
of the germ may check development of some organ at an. 
ancestral stage, while the remainder gains full development. 
Such a result, while due to atavism, would yield no evidence 
of it. To this class of influences may be due many of the vari- 
ations in offspring which so commonly occur. 
There is a further possibility to be considered: that of a 
condition the reverse of atavism. While defects occasionally 
appear in the mature body, an excess of development also at 
times appears in certain regions. This may be a duplication, 
as in the fingers and toes, the development of some limb or 
organ to a larger size than in the parents, or the appearance 
of an exerescence which has no paternal counterpart, yet, per- 
haps, may prove of advantage to the individual. If defects 
are due, as here suggested, to deficiency of energy of develop- 
ment, or partial formation in some molecular group of the 
germ, excess may, perhaps, be due to the opposite influence, a 
superabundance of energy, or excess of molecules in the group. 
The molecular groups from which the organs, tissues or mem- 
bers of the body are supposed to be derived, may possibly vary, 
as above-said, both in energy and in formative conditions, and 
