MR. JOHN MIERS ON THE GENUS CRESCENTIA. 165 



and exstipulate ; the fruit is umbonated at the summit, the pericarp being generally very 

 fragile; the pulpy matter, which I have seen only in the dried state, is of a different, 

 somewhat pithy consistence, and, when moistened, seems composed of extremely elongated 

 cells, apparently solid, flexible, and almost fibrous, intermixed with a few spiral nourishing- 

 vessels, and not formed into distinct cells around the seeds. The seeds are many times 

 as large, suborbicular, and but little compressed, with very convex laces deeply grooved 

 along the middle, thus rendering them almost 2-lobed, and closely invested by two dark 

 membranaceous integuments, the inner one in no degree contracted at the base : the 

 hilum is not basal, but is seen upon one face above the lower sinus ; the embryo is 

 very thick and fleshy, very convex on both faces, deeply grooved down the middle, 

 becomes very black in drying, the short radicle being entirely hidden by the auricular 

 lobes of the cotyledons. 



We have yet much to learn concerning the structure and growth of the ovary in this 

 genus, especially in the species belonging to the section Enallagma, the fruit of which 

 has not yet been examined in the living state. I believe, from all I have seen, that in JEn- 

 crescentia the ovary is invariably 1-locular, with four parietal seed-producing placenta?. In 

 all the fruits I have examined, belonging to both sections, they constantly show four ad- 

 nate parietal nervures, of which perhaps only two are sometimes placentilerous and seed- 

 producing in Enallagma ; but whether this be a constant or only a fleeting character, I have 

 not the means of ascertaining. The only fact on record regarding the structure of the 

 ovary in this latter section is that in Crescentia obovata, as described and figured 

 by Mr. Bentham, it is 1-locular, with two slightly prominent parietal lines of ovuli- 

 gerous placentation, but there are at the same time two other alternate prominent lines 

 without ovules. In a very young ovary of a flower in bud of the same species, from 

 Jamaica, I found it most decidedly 2-locular, with ovules attached to each side of the 

 dissepiment. May it not be suspected from these observations that the ovary is at first 

 2-celled in the very young bud, but becomes unilocular by the time of the fall of the 

 corolla ? In support of this supposition I may add that in another ovary, examined after 

 the fall of the flower (of C. cuspidata, from St. Vincent), I found it unilocular, with two 

 opposite parietal ovuligerous plaeentse, as figured by Mr. Bentham. It may also here be 

 noticed that in a longitudinal section of a fruit, not quite matured, of Crescentia elongata, 

 only half of which is glued to the paper, it is clearly shown that four distinct longitu- 



dinal cruciately disposed nervures exist, three of which remain, one having been removed 

 with the other half of the fruit ; of these, the two lateral ones are barren, but the inter- 

 mediate broader line produces seeds imbedded in laminar expansions that emanate from 

 it on both sides, which expansions are quite free from the lining of the shell at all parts 

 except along that particular line of nervure. These are all the observations that the 

 materials at my command have enabled me to make, owing to all the specimens examined 

 being glued to the paper. I state them, however, with the view of inducing others to 

 follow up an investigation worthy of further inquiry ; the facts, though inconclusive, are 

 certainly important, and require confirmation before we can attempt to draw any general 

 inferences from them. 



It would tend much to advance our knowledge of this interesting group of plants, if 



