136 DR. J. D. HOOKER ON THE GROWTH AND COMPOSITION 
tissue, and causing that fragility of that part of the ovarium which led to Griffith’s sup- 
posing that the ovarian cells were formed subsequently to the freely developed ovules. The 
true stigmatic canals I find terminating in the papillose fimbriated teeth of the annulus. 
4. Towards the period of expansion of the flower (fig. 6), a considerable change is found — 
to have taken place in the relative proportions and positions of the ovary and perianth. 
Owing to the upward direction of the development of the ovary, it is now half-superior, 
in the usual acceptation of that term. The stamens (fig. 9) form a closely imbricating 
whorl, appressed to the surface of the annulus, the stigma-like head of the central column 
appearing in the space between their apices, and the five cellular fringed teeth of the 
annulus also projecting. 
On removing the stamina, the protuberant surface of the annulus is more strongly 
5-ridged, and the groove round its orifice is deeper. A double vertical section shows 
that the column is much elongated, and its surfade covered with a more evident euticle, 
as are its capitate stigma-like apex and the adjacent walls of the annulus. The external 
surface of the annulus is also covered with a similar cuticle, except at the groove sur- 
rounding its orifice, and on the fimbriated teeth, which are manifestly the true stigmata, 
and are formed of loos papillose tissue (fig. 13). The ovarian cavities are at this period 
found to be considerably displaced, some being carried a little upwards and others down- - 
wards, so that two or three are often exposed at one vertical cut. Unrollable spiral 
vessels are now found in the vascular cord that passes downwards from the ovules, but 
. none in the transparent line that passes upwards from the ovules parallel to the inner 
walls of the annulus, these lines being formed of loose conducting tissue, terminating in 
the stigmatic teeth at the orifice of the annulus. A transverse section of the annulus 
(fig. 10) now shows five pale spots of loose cellular tissue; these are the stigmatic canals, 
placed very close to the inner walls of the annulus. 
5. At the period of expansion (fig. 7) the stamens are found to have risen from the 
surface of the annulus, and the stigmatic teeth are sphacelated and covered with pollen- 
grains, which adhere to its surface and to the surface of the mouth of the tube; but I 
have not succeeded in tracing the pollen-tubes into the stigmatic canals, probably owing 
to the excessive minuteness of the pollen in this plant. 
The differences between my own and Mr. Griffith’s observations are, first, in the ovules 
being developed in cavities of the ovarium, and not free at the base of the conical central 
body, which he has mistaken for a placental development ; secondly, in the presence of a 
cuticle over the whole surface of the said body and on the contiguous walls of the 
annulus; thirdly, in the absence of any vascular bundles passing from the ovules upwards 
into the annulus; and fourthly, in the absence of any stigmatic puncta at the base of the 
central column, and of stigmatic lines passing from these inwards to the ovules. The 
- groove round the mouth of the annulus, the absence of cuticle over it, and the structure 
and consequently the stigmatic nature of the teeth of the annulus, seem further to have 
escaped Mr. Griffith’s notice. 
If my observations are correct, the composition of this very remarkable ovarium is — . 
reducible to the known modifications of carpellary structure, and may be regarded UN 
consisting of a whorl of five connate carpels adnate to a central fleshy axis or torus; 
