384 Mr. J. Miers on the comparative Carpical 



of the cell, where the ovules became ensconced ; the parietal 

 emanations tapered upwards, gradually closing and joining- 

 together so as to form the style and stigma ; and while the cells 

 thus produced continue to grow downwards, the anatropous 

 ovules remain attached to their original placentary supports, 

 with their micropyles pointing upwards. At this stage, M. 

 Baillon's description ceases, and he does not relate what occurs 

 in the further development of Cordia, or what takes place in 

 Seliotropium or Ehretia. The differences in these subsequent 

 developments may, however, be summarized in the following 

 manner. 



In Cordia the four parietal emanations gradually approach 

 each other, to form four equal partitions united in the axis, so 

 that, at the period of flowering, the ovary is completely 4-celled, 

 with a single subanatropous ovule in each cell, attached to the 

 internal angle by a point below its apex or by its middle : the 

 completion of this growth results in a drupaceous fruit, with a 

 4-celled osseous nut, each cell with a single seed suspended 

 from below the summit, with a descending raphe terminating 

 in the basal chalaza, the radicle of the exalbuminous seed 

 being superior, and the cotyledons singularly plicated length- 

 wise. In the base and centre of the nut a large hollow is seen 

 filled with placentary tissue, from which four sets of nourish- 

 ing vessels issue, penetrating through a minute perforation 

 near the summit of each cell, and terminating in the hilar 

 attachment of the seeds. These are the very peculiar trans- 

 formations that serve to distinguish the Cordiaceo3 from all 

 other developments of the same alliance. 



In Ehretia, the ovary at an early stage is developed much 

 after the manner of Cordia : the rudiments of four ovules 

 emanate at the same points, and we see four similar inter- 

 mediate parietal enlargements ; but the basal placentary ridges 

 combine to form a compressed elevated line, running from 

 front to back across the axis, which continues to grow upwards, 

 carrying the ovules with it, or, what amounts to the same thing, 

 the principal growth of the whole takes place downwards ; and 

 in this manner the placentary columella is produced, which 

 M. Baillon does not seem to have noticed. On the other hand, 

 the parietal enlargements do not meet round the axis, as in 

 Cordia, but are thrust aside after a while ; the sinister and 

 dexter emanations form semisepta, which, on approaching the 

 columella, become suddenly reflected both ways, in parallel 

 directions, to meet the corresponding parietal groAvths from the 

 anterior and posterior walls : the consequence is that, at the 

 period of the perfection of the flower, we see two bilocular 

 carpels, each cell having a suspended ovule, while a vacant 



