2:4 NATURAL HISTOUY OF PLANTS. 



in£^ to false interseminal septa. — This character was formerly used 

 to separate genera ; it can be so used no longer. We have passed 

 the day when the JJnonas were all supposed to present these con- 

 tractions, then thought to be always absent in Uvaria ; for in certain 

 of the latter there are more evident marks of dissepiments than in 

 some of the former, and these again may have their fruit with 

 the surface quite smooth and " continuous." The genus HabzeUa 

 (A. DC), that was said to be distinguished' by its fruits, " here and 

 there irregularly swollen," like those U/ionas that have irregularly 

 moniliform berries, is at present by all included in Xylopia ; and 

 hardly is tliis cliaracter thought sufficient to characterize sub-genera 

 in certain genera. 



10. The aril. — As we know the origin of the true aril in Anonacea, 

 we can conceive a priori how this organ, formerly thought so very 

 important, can have no real taxonomic value. The soft layer sur- 

 rounding the coriaceous seed-coat, which is thickened throughout 

 in Magnoliacea, in Anonacece only undergoes this thickening around 

 or over the hilum or micropyle, in their interval, or on the sides of 

 the seed. This sort of hypertrophy may even escape notice on a 

 superficial observation, especially in the dry seed, and when it is 

 limited to a small cord bounding the two former regions. The seed 

 has then been described as wanting an aril, though this organ is 

 still represented, for its form and size can have no absolute value. 

 Never have we thought it possible to give a generic value to the 

 character of the presence of the aril." 



13. The glandular dots scattered over the surface of the leaves 

 and some other organs. — This fact and its results, as regards the 

 aromatic properties of the Anonacecs, appear to possess some im- 

 portance in certain genera ; for some consist w^holly of inodorous 

 species without dots. But here again we have nothing absolute, for 

 in so natural a genus as Anona some species are dotted and others 

 are not. 



14. The inflorescence. — It is, I think, no longer possible to found 

 genera in Anonacece on the situation and grouping of the flowers. 



A. DC, Mvm., 9. oftcu used by man." Here it would seem he 



"^ This was not the opinion of A. de Can- confuses the aril and the pericarp itself. Ae- 



DOLLK {M('m., 8), who moreover stated that the cording to the same author (1, 3), at the time 



aril, " when present, secretes an apparently resi- he wrote, no Asiatic Anonad with a clearly arillate 



nous aromatic substance at the base of the seeds, seed was known. 



