* 
90 GENTIANACEJE 
polypetalous form in a monopetalous order, than its being a 
non-adherent form in a usually adherent order. 
The fact of the opposition of the lobes of the stigma in 
Parnassia, although of less value, still is deserving of more 
consideration. It depends entirely on the assumption of 
the stigma being the denuded apex of the midrib. I say an 
assumption, because the idea seems to me totall wanting 
in some instances. 
Long ago the observation of a monstrous variety of 
Melilotus rendered it evident to me, that the stigmata form 
in no case a single organ, if carpellary leaves are continua- 
tions of the placentary margins.* The proof derived from a 
monstrosity is however unnecessary, for instances are not 
wanting, in which there are two stigmata obviously continuous 
with or opposite to the placental margins, some species of 
Phyllanthus, Nymphzea, and all filiform stigmata longitudinal- 
ly sulcate, shew this fact to demonstration. 
It follows therefore that all stigmata, atleast the dis- 
tinct portions, are opposed to the placente. The question 
then arises, from what cause does their opposition as a 
whole originate. This is easily answered, parts of a vege- 
table may cohere either with contiguous parts of the same 
organ, or with contiguous portions of two different though 
similar organs. Let us take Parnassia for instance, we have 
A stigmata, and 4 placente, hence the number of stigmata 
is 8. If the union of these take place in sucha manner, 
that the two stigmata of one carpellary leaf cohere mutually, 
then the stigma will probably appear to be a termination of 
the midrib. But if two stigmata of two contiguous car- 
pellary leaves combine, we shall obviously have one opposed 
to the placenta. And that such is the case there can be no 
doubt ; the fact of the almost universal communication be- 
tween the stigma and placenta cannot be explained by the 
f en ee ; neither need we have recourse to the destruction 
* Notule, Part I., p. 128, 
SAP SEES 
