36 oy THE FEEriLIZATION
i r om tlie subjacent tissuss in the form of a delicate membranous sac containing a stratum of
clear gelatinous material. The cells oi the e i ternal epidermis boneatli it are now no lon=-or
columnar, but broadly flask-sbaped, and open oitornally by wide circular orifices (Plate V,
% . 19), markings corresponding to which are traceablc on the inner surface ot the
detached cuticle.
The cells still adhere laterally to one another wi t h considerable tenacity, and largo shoots
of them can therctore be readily detached. This is not so with the cells of the hypodormis,
w i n c h are degenerate, softened, and so much loosened from one another as to form an
incoherent golatinons stratum (Plate V, fig. 20), saTe where the under portions of the deepest
ones adher e to the outer surface of the aohene to form a pseudo-cnticular coat ing to i t (Plate V,
fig. 18). The achenes measure about 102 x ll-r mm., and are ot a bright yellow colour.
Then- walls have a total thickness ot 0-0S9 mm. Beneath the p.eudo.cuticular coat is a
s t r a t um built up of masses of what appear to bo Tory small, completel y sclerosed cells, arranged
m columnar groups, and which represent the ultimate product of the stratum of small cells
i m m e d i a t e l y external to the internal epidermis of the immature ovary (Plate V, fig. 18).
T h i s layer is about 0-03 mm. i n thickness. Infernal to it is one about O Ooli mm. in thicknessj
a n d consisting of the modified internal epidermis. The cell cavities arc greatly reduced in
size, and are represented by a mere system ot curiously ramified, slender channels (Plate V,
fig. 18). So complete and uniform has the sclero.iis around those been, that the sites
c o r r e s p o n d i n g with the original cell-walls are onl y indicated here and there by the apposition
of the somewhat dilated extremities of the fine lateral twigs of the large oblique or vertical
channels.
T h e stigma and the distal portion of the style remain persistent and unaltered, but the
basal portion of the latter ultimately undergoes changes parallel to those taking place in the
superficial strata of the ovary. As a result o£ these, the -euticlo becomes detached as a sort
of tubular diverticulum of the ovarian cuticular sac, and the deeper tissues gelatinise, so that
It becomes difiicult without special care to procure specimens of the fiowers with the styles
s t d i adherent. •
T h e mature seeds are somewhat difficult to remove intact from the interior of the
achenes, due to the small size and resistent coats of the latter. Soaking in sulphuric acid,
however, facilitates the process, as, under the influence of this, the achenes tend to separate
more or less completely into two lateral iialves and allow the seeds to escape. They are
provided with a thin outer coat, consisting o£ empt y flattened cells in several strata, corresponding
to those of tho secundine and nucellar epidermi s o£ a n earlier period of development.
W i t h i n th^s are two strata of large cells crowded with oil globules and representing the
m a t u r e stage of the nucellar parenchyma. The embryo is of relatively large size, and is
somewhat curved upon itself, so as to leave a small space on tho funicular aspect of the seedc
a v i t y unoccupied, in which apparently a iifctle true endosperm is present. The short radicle
IS directed to the apex, and the large cotyledons to the base of the seed.
D—Characters of female flowers in mature receptacles which have not been
entered by inseetn.
I n the majority of cases the flowers retain the characters of ihose in receptacles
a t the stage for the access of insects, or at all events merely show modifications dependent
OF FICUS ROXBUEGHIt.
on dessieation. That this is so comes out very dearl y from the fol lowing measurement s of a
flower at this stage:—
Total height from base of padieelto summit of ovary . . . , 1'368 mm.
Height from base of pedicel to origiu of perianth 0-42 ram
Height from origiu of perianth to summit of ovary . • . . , . 0-94 mm
Height from uppor edge of stylar insertiou to summit of ovary , . 0'17 mm.
Breadth of ovary at level of upper edge of stykrinsertioa . . . 0-42 mm.
Leugth ot stylo and stigma
Diameters of ovule ' ' " «.oq ^
The only index to the occurrence of continued growth in this case lay in the fact tiiat the
tips of the lobes of the perianth were somewhat lower in respect to the summit of the ovary
t h a n they normally are at tho period of insect-access. Otherwise the flower appeared merelv
to have dried up. ^
I n certain cases, however, as previously mentioned in connection with the characters of
t h e receptacles, general enlargement of the flowers occurs, and certain of them even form
achenes. The following are the measurements of a tall achene-bearing flower :—
Total height from base of pedicel to summit of ovary . . .
Height from base of pedicel to origin of perianth
Height from origin of perianth to bifurcation of fibro-va
bundle
Heiglit tbeuee to base of ovary
Htiight of ovary
Breadth of ovaij at level of upper edge of stylar msertion . .
0-t>8 t
M 9 r
1-08 E
I t must be noted that this flower was one of those in which the stigma has the abnormal
t r u n c a t e form, and that, as is the rule i n such cases, the broadest part of the ovary was not
s i t u a t e d at the level of the upper edge of the stylar insertion, but at some distance above it •
t h e style being inserted lower down than iu the normal variety of flower. '
T h e achenes in size iind outward appearance are j
. , - - . -J I'te those in receptaclcs to
which nisects have gamed access, but the outer strata of the ovary are net softened
t h i s being, as already mentioned, probably due to the fact that they have not been macerated
b y receptacalar flmd. On closer examination the resemblance of the achenes to normal ones i,
found to be only superlieiah Even as regards their walls, the degree of sclerosis U
v e r y imperfect, the cell.cavities of the internal epidermis remaining relatively large and
t h e i r lateral branches being proportionately short. It is in their contents, however, . L t tho
great difference lies, as these show no traces of an embryo, and consist merely of a great thin
walled empty sac representing the dilated nucellus and secundine. In many cases all the
cells in its walls are thin, flattened, greatly extended superficially, and almost or quite empty
I n a few insianoes a feeble at tempt at aecnmnlation of albumen has seemingly occurred the'
cells corresponding with tho nucellar patcnchyma showing a certain number of pale globules
within them. The devolopmont in such oases forms a sort of parallel to the imper fect evolu.
t i o n of tire mal e flowers which, as has been already shown, sometimes occurs "
a p a r t from the access of insects.