1906] CHAMBERLAIN—OVULE OF DIOON 333 
crowded and interlaced and sometimes even branch, thus giving rise 
to an extremely tenacious tissue, 
The inner fleshy layer—This layer contains the inner vascular 
system, but otherwise consists of rather uniform, slightly elongated 
parenchyma cells. The cells are smaller than those of the main 
body of the nucellus, but there is no definite boundary between them. 
The boundary between the inner fleshy layer and the stony layer is 
less indefinite, and as early as January the two layers may be peeled 
apart, though with a rather uneven break. In small November 
ovules the inner fleshy layer has about the same thickness as the stony 
layer, but from this time the fleshy layer grows more rapidly, and in 
early December ovules is much thicker than the stony layer (fig. 10). 
The rapidly growing endosperm then begins to encroach upon the 
adjacent cells, which we regard as nucellar tissue intimately united 
with the inner fleshy layer of the integument, although it must be 
admitted that the ontogeny shows no indication of such a union. 
The encroachment continues until in the ripe seed all the tissues 
between the endosperm and the stony layer is reduced to a thin dry 
membrane, which peels off easily and cleanly, and shows very clearly 
the distribution of the inner vascular system. The inner fleshy layer 
is to be regarded as belonging to the integument rather than to the 
nucellus, because it is continuous with the inner fleshy layer of the 
free portion of the integument; because the bundles which it con- 
tains often extend into the inner fleshy layer of the free portion of the 
integument but never into the free portion of the nucellus; because in 
half-ripe ovules the fused portion of the nucellus may be peeled away 
from the layer containing the inner bundles; and, still more impor- 
tant, because such a conclusion is warranted by a comparison of the 
cycad ovule with those of fossil gymnosperms. 
The phylogeny of the three layers.—This is an interesting question. 
Has the single complex integument always been single, or does it 
Tepresent two integuments which have become fused? This is a 
question to which no definite answer can be given, because the early 
development of the ovule has never received sufficiently careful study 
in any cycad. The early stages in the development of the integument 
are passed before the cone breaks through the scale leaves, and conse- 
quently no material is likely to be secured in greenhouses. In the 
