and on various Plants related to them. 321 



the anthers suggests the probability that the mechanical raeans for promoting 

 dehiscence are confined to the middle, not as they v^ery generally are, to the 

 inner tissue. 



Neither Mr. Harvey nor M. Endlicher has noticed the remarkable struc- 

 ture of the stigma, either as regards its striate appearance, arising from its 

 lamellate composition, or the very evident definite grouping of the lamellae. 

 Possibly in the state of nature both of these are concealed by stigmatic secre- 

 tion, or by a particular state of the tissue disappearing on maceration. The 

 apparent opposition of its lobes to the lobes of the staminal column does not 

 appear to have been noticed. 



The observations I have made on the placentation do not entirely agree with 

 those of M. Endlicher, who has described the placentae as being parietal in the 

 unfecundated state of the ovarium*. In my specimens, which embrace a con- 

 siderable range of development, they have always appeared to me to be free 

 and pendulous, bearing ovula over their whole surface ; and this agrees with 

 the observations of Mr. Harvey. The determination of this is of some import- 

 ance, since if the placentae are free and pendulous throughout, another objec- 

 tion appears to me presented to the placentary hypothesis of M. Schleiden. 



The antitropous nature of the ovula, although sufficiently obvious in the 

 earlier stages of their development, soon ceases to be discernible even under 

 pressure. 



Obs. ni. — So far as my experience goes, the vegetable kingdom does not 

 present a more complex or anomalous instance of the structure of the pistil- 

 lum. Considered as an instance of multicarpellary structure, the stigma ap- 

 pears to admit of satisfactory explanation, and to be analogous to the stigmata 

 of Papaver and Nymphcea ; the space between each lamella corresponding with 

 a carpellary leaf, and each lamella itself being compound, as almost always 

 happens in such instances. This would be in my opinion the obvious struc- 

 ture, should M. Endlicher's observations regarding the placentae of the un- 

 impregnated ovarium being parietal prove to be correctf. But the evident 



* Genera Plantarum, p. 75, in the observation. 



t It may be gathered, perhaps, from Mr. Brown's remarks on the occasional limitation of the pla- 

 centa to the apex of the cell of the carpellary leaf, that he thinks it probable that Hydnora is multi- 

 carpellary. — See 'Plantae Javanicse Rariores,' part ii. p. 109. 



2 u 2 



