46 WorsdelL—The Structure of the 
reduced to a scale. He says also that in Thuja , Cupressus , 
Juniperus , the scales bearing the ovules are modified leaves 
of the primary axis, as there is no sign of any other bract, 
and in the vegetative shoot there is a gradual modification of 
the leaves from the seedling upwards. 
Alexander Braun, the renowned German botanist, must 
take credit for being the first, viz. in 1853, to promulgate 
the view that the seminiferous scale consists of the two first 
leaves of a bud axillary to the bract , these leaves being fused 
by their margins (Fig. 3). 
Baillon (54, 62), on three separate occasions, in important 
papers, advocates the ovarian theory of the ovule, making 
a determined stand against the opposite theory of Gymno- 
spermy. He bases the chief 
weight of his argument on the 
development of the organs con¬ 
cerned ; he observes the semini¬ 
ferous scale appearing as a pa¬ 
pilla in the axil of the bract, 
just like the bud of an incipient 
branch (Fig. 4), and concludes 
therefrom that it must be of 
axial nature, forgetting the fact 
that structures may, even in 
their youngest stages, be already 
congenitally modified, i.e. inheriting directly, without passing 
through the previous progressive stages, the adapted structure 
of the same organs in their parents. His firm conviction that 
the integument of the ovule is an ovarian wall is based chiefly 
on the fact that, in the earliest stage of development, that 
structure appears as two distinct papillae, which, as he thinks, 
must be the two carpels composing the ovary. But one fails 
to see exactly why this should be the interpretation, for there 
is variation in this respect among plants. Assuming the 
ovular envelope in the genus Podocarpus to be an ovary, we 
find that in one species, P. chinensis , Sw., it arises as a con¬ 
tinuous ring, but in another species, P. dacrydioides , A. Rich., 
Fig. 4. Diagram of part of cone of 
Abietineae, showing the origin of 
the seminiferous scale («) in the 
axil of the bract ip ); ap = apex 
of axis of cone. (After 0 rsted.) 
