Mr. J. Miers on the Bignoniacese, 259 



placentae ; but a distinction is manifested in the subsequent de- 

 velopment ; so that, in examining the fruit, we must bear in mind 

 the previous structure of the ovary. The want of materials has 

 prevented me from investigating this subject. I have been able 

 to examine only a single ovary of Crescentia, which was partly 

 injured by caries ; but this satisfied me that it had only two 

 parietal placentae. Kigelia I found similarly constructed, and not 

 bilocular, with ovules borne on the centre of the dissepiment, as 

 is represented in Delessert^s ' Icones/ v. tab. 93 b. fig. 3 : the 

 appearance there shown is the result of the touching of the op- 

 posite projecting placeutse, which, in the younger state of the 

 ovary, and even after the fall of the corolla, I have found sepa- 

 rated by a long interval. In Schlegelia I have also verified the 

 same structure. In regard to the fruit of Crescentia, the details 

 of Gaertner are precise, are illustrated by good figures *, and 

 appear worthy of full confidence : it is circular in its transverse 

 section ; its indehiscent shell, though thin in substance, is hard 

 and somewhat ligneous, marked externally and internally by 

 four equidistant longitudinal ridges, the cavity being filled with 

 a soft pulp, in which the seeds are imbedded. The description 

 of Gardner t, in regard to the fruit, is similar ; but he gives a 

 very different account of the ovary, which he says is " 1 -celled, 

 with four fleshy parietal polyspermous placentse placed one on 

 each half of the pericarpial leaves, and at equal distances from 

 each other.^' There appears some error in this statement ; for it 

 is contradicted in his account of the fruit, which states, " peri- 

 carp woody, consisting of two indehiscent carpels placed anterior 

 and posterior to the axis of inflorescence." Of the existence of 

 two opposite placentse there can be no doubt ; the two inter- 

 vening prominent lines, in the case which I observed, were bare 

 of ovules, and seemed to arise from the line of junction of the 

 thickened sterile margins of the normal carpels, similar to what 

 I observed in the ovaria of Kigelia and Schlegelia : at first sight 

 these seem to have four lines of placentation ; but a more ca eful 

 observation shows the presence of two only. If this view of the 

 structure of the ovary in the Crescentiacece be correct, it will be 

 represented as in fig. 16, that is to say, of two carpels pi„. i5_ 

 placed face to face, which are placentiferous on their 

 midribs and conjoined by their sterile margins, a 

 structure that will be seen to correspond with the 

 Eccremocarpece (fig. 14), differing only in the greater 

 thickening of the margins of the carpels. It remains to 

 be ascertained whether the pulp of the fruit in these genera results 

 from a secretion formed at the internal surface of the ovary, or 



* De Fruct. iii. 230. tab. 223. f Hook. Journ. Bot. ii. 423. 



