GENERAL REMARKS. 589 
similar to those found on the back of the style sparingly and 
in immense numbers on the tube of the calyx. 
These spaces, do not communicate with the placente. 
The stigmatic canal has the usual structure: it is lined 
with stigmatic tissue and presents 3 faces, and 3 round angles, 
which angles correspond with the fissures between the separ- 
ablé lobes. 
The styles are very vascular possessing many fascicles, the 
dilated part presents two converging and diverging from the 
upper part of the style. 
I believe in fact, that the sepárable lobes of the stigma are 
in reality compound, that each style is bifid and each divi-' 
sion expanded into a lamina curved outwards, and then in- 
wards, These divisions mutually cohere, simulating a ae 
of a true stigma, 
For the lines of communication with the stigmatic canal 
alternate with these lobes, the lobes themselves have no 
. Central vascular axis which they would have, did they cor- 
.. respond with the midrib of the leaf. 
It is curious that if such explanation be correct, the nature 
of the inflexion of the style, in its expanded part will corres- 
pond in a considerable measure with that of the carpellary 
leaf, as seen in the very young states of Coccinia indica. 
It follows that if this be the case Dr. Wights hypothesis 
Cannot be correct, for that necessarily requires that the midrib 
‘shall correspond with the axis of the placentze. 
— ]t is nota little remarkable that the apparent inversion 
.. 9f the carpellary leaf should be so continued in the dila 
-Part of the style. 
My ideas of the stigmata of this plant (the kunkree) are re- 
presented i in the annexed diagram, Pl. DCXVII. Fig. II. a, 
4, a midrib, b ditto, c direction of the line of convolution, 
style with vessels, e stigma. ` 
- In Gardenia there is an evident tendency to dilatation of 
the Styles, which are in their upper part battledore-shaped, 
ley are slightly bilobed, the stigmatic tissue occupies the 
