506 The American Naturalist. [June, 
and independently in both, as soon as certain cells of coherent 
groups become over-nourished and incapable of further seg- 
mentation unless brought into contact and fused with the mi- 
nute male elements, or one which, as we have seen, is the pro- 
duct of an exalted segmentational power which is transferred 
to the female element in the act of fertilization. Both kinds 
of sexual products were probably at first, and still continue to 
be, dehisced from the parent organisms as useless products of 
over-nutrition, after further recapitulative growth in the form 
of new axes or of individuals, growing in organic union, as in 
colonial organisms, became impossible, due to crowding, the 
culmmation of seasonal growth or the morphological speciali- 
zation leading to definite or constant formal individuality. 
All the facts which I have been able to gather lead to the 
conclusion that there is a relation between the difference in 
size of the male and female elements as to the number and ra- 
pidity of the subsequent segmentations of the resulting 
oosperm or oosphere. If the elements are alike there will be 
comparatively few segmentations ; if greatly unlike, many suc- 
cessive segmentations seem possible. 
The foregoing hypothesis affords clews to the reasons for 
variations in the fertility of species, the origin of viviparity 
and placentation, the infertility of irrelated forms, the origin of 
tood yolk in ova and of pelagic eggs, the evolution of primary 
and secondary sexual characters, the interrelations of plants 
and animals, and a consistent and simple theory of inheritance, 
which is in harmony with all the facts of reproduction in plants 
and animals. 
This hypothesis also discloses some of the apparent reasons 
why there is so frequently a great difference in the size of the 
sexes, as in fishes, where the male is smallest, and especially 
in those arthropodous forms, in which the males are micro- 
scopic and attached to or parasitic on the females, as in some 
Copepoda and Cirripedia. The extraordinary feeding and 
nursing habits of social Hymenoptera, efficient in determining 
the sex or neutrality of offspring, also acquire a new signifi- 
The first steps by which the over-growth of the sexual ele- 
merits through over-nourishment is seen in the most primitive 
oi ail known non-parasitic, free-swimming, multicellular forms, 
namely, Volvox. Its life history proves that the multicellular 
condition can be, and probably was, attained directly by the 
over-growth and subsequent segmentation of a single cell in 
