IN THE CLUSIACE.E, MAGNOLIACE^, ETC. 83 



and its connexion with the raphe, in the relative position in which the latter is found, 

 have evidently escaped the observation of these eminent botanists, as they had previously 

 done that of Gsertncr ; it is, however, imjiortant to attend to these distinctions, as they 

 form essential and convincing elements in this inquiry. 



Still more recently a paper from Dr. Asa Gray has appeared in ' Hooker's Kew Journal 

 of Botany' (vol. vii. p. 243), " On the Development and Structui'e of the Integument of 

 of the Seed of Mciffnolia." This was written in reply to my observations on the aiilliform 

 nature of the outer tunic, to which I have above referred. In that note, this accomplished 

 botanist warmly defends his former opinion, which he maintains with great ingenuity and 

 candour, but I find nothing there stated that disproves the conclusion of most botanists, 

 regarding the natm-e of the several seminal tunics in question, which I have here 

 endeavom-ed to confirm. If I had not been convinced that the issue admits of the most 

 simple and demonstrative proof, I should not have presumed to contest the united 

 authority of the distinguished American professor, and of the highly talented authors of 

 the ' Flora Indica,' whose opinions are entitled to especial consideration. 



The argtmaent that the scarlet external envelope cannot be an aril, because the latter 

 " must needs have an opening at the top," stands upon very feeble ground. This was the 

 definition of St. Hilaire, who classed the different forms of arilliform coatings under two 

 denominations : 1. the true aril where the envelope has an opening in the top ; 2. the 

 false arU, where the coating that siu-rounds the testa is entire. Dr. Planchon, however, 

 who ably defended and extended the views of his friend, in regard to the nature and 

 origin of the aril, has shown that this definition does not hold good, and among other 

 cases, he cites a species of Tetracera fi-om Java, where the perfectly true arU forms an 

 entire tunic. I need hardly refer to Gsertner, who also divides the arU into two kinds, 

 the complete and the incomplete ; notwithstanding that in some of the instances of the 

 first kind, he has mistaken the endocarp for an aril ; others, which he enumerates, possess 

 a complete ai'iHiform covering, among them Nephelium, which in this respect offers a 

 good analogy with Magnolia. Another instance is stUl more remarkable, because it is 

 recorded by Dr. A. Gray himself, in the same admirable work, and within a few pages of 

 his description of IlagnoUa : it occurs among the Anonacece, figured ia the genus Asimina, 

 (plate 27) and described (p. 57) as entii-ely covering the testa, without the mention of any 

 aperture in the apex. There is indeed no reason why an aril, which is a development 

 from the j^lacenta, may not form an entire coating, as well as the true integuments of 

 the seed, originally cup-shaped processes. The argument of the inadmissibility of an aril, 

 because it is not open at the top, therefore completely fails. 



In regard to the remaining arguments of Dr. A. Gray, I can only repeat what has just 

 been demonstrated, that the nourishing vessels fi-om the placenta, as he has figm-ed them, 

 only penetrate the primine of the o^iilc of Magnolia at the gangylode* ; they cannot and 

 do not enter at the opposite extremity ; but if we examine the ripe seed, we find the 



• For the sake of demonstration I have here given the name of gangylode to that common point of imion of the 

 primine, secundine and terciue in the original base, now the summit of the inverted ovule, which afterwards form the 

 diapjle of the testa, and the chalaza of the imier integument of the seed. 



M 2 



