144 
The peduncle grows rapidly and becomes several times longer 
than the sporocarp itself—(Fig 5). The growth upon the upper 
side of the latter is stronger than upon the lower side, and in con- 
sequnce, it becomes bent over, nearly or quite at right angles to 
the peduncle. With the enlargement of the sporocarp the cavi-: 
ties within become deeper and wider in a direction parallel to the 
broad surfaces of the lobes; but the radial growth of these keeps 
pace with the longitudinal growth, so that the space between the 
inner surface of the lobe and the columella, is very narrow. The 
growth is especially active in the inner epidermal cells, which pro- 
ject more or less and form a cushion running vertically along the 
median line of each lobe. This cushion is the sorus (Fig. 5.6.8. ), 
and as its surface cells develop into the sporangia, it nearly fills 
the cavity in which it lies. A tranverse section at this time shows 
that the portions separating the cavities are composed of about 
four layers of cells. 
The fibro-vascular bundle which traverses the center of the pe- 
duncle divides into four branches at the base of the sporocarp, and — 
one of these goes to each lobe and forms a sort of midrib (Fig. 
6.). Later each of these bundles sends off two lateral branches 
that follow the margins of the lobe. A cross section of a sporo- 
carp at this stage (see Fig. 9), shows these as groups of small cells 
at nearly equal distances from each other, one at the center of the 
lobe, the others close to its margin. By this time the epidermal 
cells of the outside of the sporocarp begin to thicken. This is the 
first indication of the hard shell found in the ripe fruit. 
While these changes are going on in the outer tissues of the 
sporocarp, the sporangia have begun to develop from the surface 
cells of the sori. -Active growth begins in these cells, which be- 
come elevated as papillze above the surface of the sorus. This is 
most marked in the basal or older cells, but proceeds rapidly to- 
ward the upper end of the sorus. While in a general way we may 
say that the sporangia arise in acropetal succession, still 
new ones arise later among the earlier formed ones, without refer- 
ence to their position; indeed all of the surface cells of the sori 
may be regarded as potentially, at least, sporangium mother-cells. 
There is a good deal of variation in Pi/u/aria, as in many ferns, 
as to the direction of the first division walls in the young sporan- : s 
