54 WorsdelL—The Structure of the 
arises immediately in the axil of the bracts, as in other 
Cupressineae; later on small scale-like outgrowths appear 
between the ovaries and the bracts, which cannot be carpels 
as they appear later than the so-called ovules. (2) The 
ovaries are often not placed in the axils of these inner scales, 
but alternate with them, and are lateral. (3) The number of 
the ovaries is independent of the number of these scales ; often 
more scales occur than ovaries, or vice versa. (4) In the 
anomalous development of the scale it may fuse with the 
ovary and increase the size of the latter. (5) In anomalous 
development of ovaries their foliar nature becomes apparent.’ 
He considers that abnormalities are of importance and 
follow a definite law. 
In Juniperus Hartwissiana he finds certain objections to 
the carpellary theory of the bract. (1) There are an indefinite 
number of ovaries at the base of the bracts, three, five, or four; 
in the last case two ovaries occur in the axils of each two 
bracts, while the third is left empty; which shows that the 
presence of ovaries is not a necessary character of the bracts. 
(2) The ovaries often alternate with the bracts, as if continuing 
the whorl of three leaves upwards. The ovary first arises as 
four small protuberances. 
These are his concluding remarks:—‘(i) The development 
of the ovule of the so-called Gymnosperms corresponds 
entirely with that of the ovary, and not the ovule, of other 
Phanerogams, viz., (a) the so-called integument arises earlier 
than the nucellus; ( h ) as a result of this, the former arises 
from the receptacle and not from the nucellus, as a true 
integument should do ; (c) it is formed of distinct carpel- 
leaves, which fuse sooner or later, a fact which does not 
occur in any integument. (2) The so-called integument arises 
independently, without a nucellus, and remains long separated 
therefrom, surrounding it loosely, and leaving often a con¬ 
siderable space between. (3) The simple structure of the 
ovary of Gymnosperms is no reason for considering the latter 
as an ovule. (4) The anatomical structure of the so-called 
integument is too complex for such an organ. (5) The 
