216 The Botanical Gazette. (Jame, 
Over the greater part of its surface, the outer integumentis — 
in contact with the ovary wall. Figs. 3, 4. 
In a considerable number of specimens, a pit was observed 
a short distance from the base of the style, on the posterior 
side. In some cases the depth of the pit at this stage was — 
equal to one-half the thickness of the ovary wall. The edges — 
were rounded and complete, showing that this space was due — 
to a failure of the parts to unite completely. In younger — 
specimens in which the embryo-sac had not yet developed, — 
this pit was generally to be found. In many of these cases — 
the pit extended entirely through the ovary wall exposingthe 
already developed outer integument. It seems that at this — 
stage of development, the edges of the carpel have not ye — 
completely united. Fig. 2. 4 
The outer integument is, in the main, four layers of cells 
in thickness. Toward the top of the ovule it is reduced to two 
layers, except at the base of the style where it is thickened j 
into a broadly wedge shaped mass which projects for a short 
distance into the base of the style, fig. 7. The short portion — 
projecting over the anterior face is terminated by 4 SS* — 
layer of wedge shaped cells. This integument is for ™ 
greater part in close contact with the ovary wall on the out 
side and with the inner integument on the inner side, 
space between it and the ovary wall. As this f 
ersed by the descending pollen tube, pea? ie 
aid in determining its course. The cells constituting this ; 
tegument are generally short-prismatic, rather regular 4 
have very thin walls, figs. 3, 4. The short wedge-‘ ; 
tion found below the micropyle is in close contact W! 
part of the inner integument, but is removed from t 
wall by a considerable space, fig. 8. 
The inner integument varies in thickness 
The cells as in the outer integument are : 
form but are considerably larger. The inner i 
closely in contact with the nucellus. At no poln rganssave 
integument i : ‘+h anv of the adjacent © 
guments strictly united with any a nate 
only at the points of their origin. At interv jon a 
may be seen between adjoining structures. isin tothe 
cribed to the integuments seems to be that of prot ie that the? 
parts enclosed. Holzner'! regards it as prope” 
