112 Dr. J. E. Stocks on the Structure of Cucurbitaceae. 



belonging evidently to a ternary whorl, tlieir binary adhesion in 

 some cases, their separation (even as to vascular bundles) in 

 others, as also the opposite twisting of their loculi in contiguous 

 members ; whereas some or other of these particulars will stand 

 in the way of other views. 



Ovary. — The ovarial leaves are sometimes two (Mukia, Pilo- 

 gyne), generally three. 



In the three-leaved ovary we have three dissepiments proceed- 

 ing from the parietes and three from the axis, which last bear the 

 ovules on their parietal extremity. 



These appearances are variously explained. 



1. Dr. Lindley supposes a valvate aestivation of the carpellary 

 leaves. According to this view, the dissepiments are spurious, 

 three proceeding from the placentae and the three intermediate 

 ones from the midrib of the carpels. 



2. According to Schleiden's views, the placental dissepiments 

 must be regarded as prolongations of the axis, extending into 

 the cavity formed by carpellary leaves with an induplicate aesti- 

 vation. This opinion was long ago (1823) taken by St. Hilaire. 



3. Dr. Wight supposes the carpellary leaves to have a redu- 

 plicate aestivation, and the cavity of the ovary to be completed on 

 one side by the calyx. The intermediate or primary dissepiments 

 would have to be regarded as spurious. 



4. Arnott (Prodromus Fl. Pen. Ind. Orient.) and Endlicher 

 (Gen. Plant.) describe the carpels of Cucurbitacece as involute. 



The evidence afforded by dissection and by analogy proves the 

 correctness of the last of these views. 



1. In many Cucurbitacece whose corolla is conical in bud, the 

 aestivation of the upper part of the corolla is beautifully involute, 

 and presents a remarkable similitude to the young ovary : so 

 that the bending of the carpellary leaves is not a forced expla- 

 nation, but is just what happens in Cucurbitacece when the floral 

 leaves meet in the axis. 



2. In those Cucurbitaceous fruits whose vessels lignify and 

 whose cells encrust, we can trace the leaf-skeleton following an 

 involute course, and in none better than in many species of Luff a. 

 Breaking off the outer shell (calyx) we come to a fibrous layer 

 which runs externally chiefly round the fruit, and internally from 

 top to bottom. This sends in processes at three points only (pri- 

 mary or barren dissepiments), which after meeting in the axis 

 turn outwards into the cavity of the ovary and bear the seeds. 

 In Luffa pentandra, just before hardening has commenced, on 

 removing the operculum the primary or barren dissepiments are 

 plainly seen to be composed of two layers when we examine them 

 at the apex of the fruit where seeds are not developed, and less 

 plainly below, owing to the pressure of the seeds which indent 



