and on various Want* rdrniod t<> tkem. ;i.il 



ample occurs id Orobanche*, as may be ascertained by the nrtfiminitkn of 



the stigma at very early periods, and of the situation of the vascular bondlct 



of the style, which are anterior and posterior, as in all the allied genera 



I have examined. A similar sort of cohesion occurs in PapaveracetB, and 



perhaps in all cases in which the stigmata, being apparently equal in num- 

 ber to the placenta 1 , are said to be opposite to them. On this point, the 

 stigmata of Ltnaria purpurea and Thmbergia aUtia (a/ha), bear with consi- 

 derable forcef. 



The stigmatic surfaces may be divided without any particular reference to 

 the state of the styles or composition of the ovarium. Of this, Acahfpha and 

 the two genera which have induced these observations are instances in ex. 

 Some species of Bragahtia, on the contrary, appear to present only three 

 styles to four carpclla. And I think it may be said, that, the stigma, being an 

 extension of or continuation from cellular surfaces, frequently of very irregu- 

 lar growth, is not to be expected to present a constantly definable form J. 



* Regarding this I can speak with such confidence as I may, when Dr. Lindley has expressed 

 himself positively to the contrary ; and has, from the consideration of Orobanchar, so extended the 

 possible origin of the placentae, that he conceives these organs to arise from no definite portion of the 

 carpellum, but to vary in origin according to specific organization. (Introd. to Botany, 2nd ed. p. 203.) 



f These instances also bear on Dr. Lindley 's supposition regarding the composition of the " inter- 

 vening web or membrane" of the stigmatic apparatus of Habenaria, Bonatea, &c. (Gen. Sp. Orchid. PL, 

 Preface, xi.) 



X In all cases in which stigmata are to be observed with referenre to the composition of the ova- 

 rium, I have, I think, derived advantage from the examination of their surfaces, their vascular supply. 

 and its connexion with that of the style and ovarium. If the stigmatic divisions have both surfaces 

 uniformly stigmatic, still more if they present no vascular fascicle, I take them to be stigmata alone. 



On the contrary, the circumstances of their outer surfaces or dorsa not being entirely stigmatic, and 

 the presence of vessels, which, so far as I have Been, have a clavate termination, applied to cases de- 

 scribed as stigma bifidum, bilamcllatum, &c, will show that these terms of division have reference to 

 the style. They will also I think show, that many of the Euphorbiacece cited by Schleiden as having 

 9tigmata only, possess bipartite styles ; that style* exist in some Graminea at least; and that in Com- 

 posites the term rami styli is more correct than that of stigma bilobum. 



In those cases in which the stigmatic surfaces are simple and really confined to the style of the same 

 carpellary leaf, the form of the stigma will generally depend in a great measure upon the extent to 

 which the convolution of that part of the carpellary leaf (generally a cuspis) which forms the style is 

 carried. If the convolution be complete, presenting an equal margin, we have a terminal stigma without 

 any sinus, as in many Leguminosce, Mirabilis, &c. If the convolution is less complete, we may have a 

 reniform stigma or one of any discoid form with at least an anterior or inferior sinus, and according as 



