458 MR. J. R. GREEN ON THE ORGANS OF 
stem. The main difference between the two parts is the much 
greater abundance of them in the ovary. As already stated the 
ducts are in several rows, and they anastomose frequently with 
each other (Pl. XL. figs. 18 & 19). In the ovary of the thin-walled 
species they arise from cells close under the epidermis, the cells 
of which are slightly smaller just over them. As the fruit 
grows, the epidermis-cells which do not overlie the passages grow 
also, keeping about the same relative size ; but those which abut 
on the secreting tissue remain small, so that a depression on 
the surface corresponds to a passage (fig. 21). In the thick- 
walled ovaries, as the wall thickens fresh passages are developed 
in it continually. Consequently a great amount of resinous 
exudation is noticeable in cutting such a fruit. There is in 
these no modification of the epidermis. In the thin-walled 
ovaries the passages may extend quite through the mesophyll 
and touch the epidermis at both surfaces ; or they may be of less 
diameter and abut on only the outer one. Though they are 
always in contact with the epidermis, the latter takes no part in 
their formation, as, according to Martinet *, it does in the similar 
glands of Dictamnus. 
In the floral leaves of the species of stouter habit there are - 
similar passages. In the sepals often several rows may be found, 
which differ in no way from those of the ovaries. The petals show 
a rather remarkable distribution. The sstivation of Hypericum 
calycinum is convolute; when the flower begins to expand the 
sepals open some little time before the petals, and leave these 
latter twisted round and covering over the essential organs of 
the flower. The centre of the whorl of sepals is therefore 
occupied by a small knob-like body of bright yellow colour and 
glaucous appearance, which is composed of the stamens and 
pistil covered over by the contorted petals. On opening out 
the separate petals, each is found to be of very different texture 
at its two lateral edges. The part which was outermost in the 
coil is fitted for its protective functions by being thick and firm, 
and having an epidermis whose cells are strongly cuticularized. 
The side which in the twisting was in contact with the stamens, 
and was therefore covered over and kept from the air by the 
thicker portion, is thin and delicate, with a somewhat papillose 
epidermis whose cells are not cuticularized. The two kinds o 
structure gradually blend one into the other at about the middle 
* Loc. cit. 
