growth of the glands continued up to the fourteenth day, after which
regression set in. This shows that the development of the milk glands in
rabbits is due to the corpora lutea. On the other hand, Lane-Claypon and
Starling state that in rabbits the corpora lutea diminish after the first
half of pregnancy, while the growth of the milk glands is many times
greater during the second half than during the first half of the period,
and during the second half the ovaries may be removed entirely without
interfering with the course of pregnancy or the normal development of the
milk glands. It is evident, therefore, that in rabbits, whatever influence
the corpora lutea may have in the first half of pregnancy, they have none
in the second half, and that at this period the essential hormone proceeds
from the developing foetus or foetal placenta. Again, if it is the
withdrawal of a hormone stimulus which changes the milk gland from growth
to secretion, it cannot be the corpora lutea which are exclusively
concerned even in _Dasyurus_, for they persist during lactation, while
secretion begins shortly after parturition.

Gustav Born suggested, and Fraenkel tested the suggestion experimentally,
that the corpus luteum of pregnancy is a gland of internal secretion whose
function is to cause the attachment of the ovum in the uterus and the
normal development of uterus and placenta. Fraenkel found that removal of
both ovaries in rabbits between the first and sixth days after
fertilisation prevented pregnancy, and that the same result followed if
the corpora lutea were merely destroyed _in situ_ by galvano-cautery.
Either process carried out between the eighth and twentieth days of
pregnancy causes abortion.

Lane-Claypon and Starling also found that removal of both ovaries in the
rabbit before the fifteenth day was apt to cause abortion, but at a later
stage the same operation could be performed without interfering with the