488 Mr. J. Miers on some genera of the Icacinaceae. 



soft texture of the ovule, and the minuteness of the parts, it is 

 difficult to extract it entire ; it bears at first sight the appearance 

 of a single ovule, but from the overlapping of the visible edges, 

 it would seem as if composed of two ovules pressed together : as 

 I failed to separate them, owing to the reasons just mentioned, I 

 cannot state with certainty that this body consists of two ovules. 

 The analogy of this structure with what I have observed in Ste- 

 monurus and other genera will be found most complete, and it 

 offers additional testimony to support what I have urged, in 

 regard to the normal structure of the whole family. 



I have alluded to the fact of the unusual occurrence of a 

 raphe-like cord, imbedded in the pulp of the fruit, originating 

 in the calyx, ascending along the external ventral side of the 

 nut, and attaching itself to the strophiole of the seed which 

 protrudes through the foramen just below the apex : this cord 

 consists of two distinct threads, and apparently results from the 

 extended remains of the two abortive cells, or it may consist 

 merely of the nourishing fibres, become thickened, that con- 

 stituted originally the axile column of the united carpels : its 

 marked appearance in such a position is certainly an occurrence 

 worthy of notice. I shall be able to show, that under certain 

 circumstances, this external cord disappears, and the phenomena 

 then observed all tend to confirm the inferences just drawn in 

 regard to the true nature of this process. 



Another fact remains to be mentioned that offers very instruc- 

 ive evidence. The two species of Pennantia from New Zealand 

 and from Norfolk Island, both produce a drupe containing a 

 solitary, hard, three-cornered nut, as has already been described : 

 though very different in dimensions, their structure is precisely 

 alike. The fruit of P. Cunning hami, on the contrary, encloses 

 a thin, soft and coriaceous putamen, which is oval, 1 -celled, and 

 contains a single seed, resembling the two preceding species in 

 structure and attachment ; but the external cord is adherent, and 

 cannot be detached from it without rupture. I have found, how- 

 ever, one instance in which the putamen is distinctly 2-celled, 

 the cells being separated by a thin dissepiment, and each filled 

 by a single seed, suspended in the usual manner from the 

 summit of a thick internal cord, which originating in the base, 

 proceeds close to the walls of the putamen, attached to one side 

 of the dissepiment, reaches the summit of the cells, where it 

 throws out two strophiolar processes, one into each cavity, from 

 which the seeds are respectively suspended : it need hardly be 

 suggested that the cord in question must be the elongated relic 

 of the abortive third cell. This is an instructive fact, showing 

 how necessary it is to trace the growth of seeds through all their 

 successive stages from their origin, in order to understand their 



