On Reproduction, Animal Life Cycles. 
91 
sexual reproduction and not, as in metagenesis, of sexual and asexual. 
It is not true heterogenesis unless there is a succession of at least 
two generating individuals. Thus the queen of the Honey Bee, as 
Francois Huber proved , 26 normally lays fertilized worker eggs, then 
unfertilized drone eggs, then fertilized worker eggs again, but this 
is not true heterogenesis because normally only queens lay eggs, and 
not workers, reproduction being accomplished by a single individual. 
Heterogenesis is well illustrated by digenic Trematodes, Rotatoria, 
and various Crustacea and Insects. It is discontinuous development 
because it embodies breaks between successive somatic individuals. 
The beginning of the cycle must be considered the fertilized egg, 
because of the primitiveness of amphigony; the termination of the 
cycle would therefore be that individual which produces eggs that 
need fertilization; these would be respectively the first and last 
stages of the cycle. The process of parthenogenesis would then be 
characteristic of stages intermediate between these. Thus while in 
metagenesis the secondary interpolation is asexual generation, in 
cases of heterogenesis it is parthenogenesis. Yon Baer’s idea of 
paedogenesis meant egg production by a larva; if instead of larva, 
which we have particularly defined, we substitute the term immature 
stage or immature individual, then we find that paedogenesis is a regu- 
lar accompaniment of heterogenesis. 
By ekdytic life cycles I have intended those in which an immature 
stage, that is not a larva, produces certain structures that are thrown 
off or moulted before maturity, and in which that immature stage 
is sedentary and enclosed. Such are the cases of viviparous animals 
where the young develop intraparentally, as well as cases of develop- 
ment within a strong egg shell whether the egg be within or without 
the parent. The conditions are the same in both, development within 
an enclosed space and consequent immobility of the embryo. In both 
also particular foetal or embryonic membranes are formed for me- 
chanical protection, respiration and excretion; thus in a viviparous 
Vertebrate as well as in an oviparous insect an amnion and serosa 
are produced. Viviparity and oviparity are states only of the parent, 
not of the embryo ; what induces the production of foetal membranes 
is not one of these conditions but growth within a closed cavity. This 
is discontinuous development because some temporary organs formed 
by the embryo, such as an amnion, serosa or allantois, are moulted 
at birth and take no part in forming the adult organism. It differs 
from metagenesis in that there is no sexual generation involved, for 
the parts thrown off do not compose a second individual and are 
structurally simpler than the embryo; and in that there is a quiescent 
28 Nouvelles observations sur les Abeilles, 1814, 2 me ed., Paris. 
