48 



ROYAL HORTICULTURAL SOCIETY. 



" But these are not the only changes which have taken place in 

 the flower and very young fruit, there is another which is not less 

 worthy of attention, and without which the fruit would remaic 

 incomplete. We have seen that at first the bud was sessile, or 

 nearly so ; gradually the rudimentary peduncle is elongated and 

 assumes the well-known form, but at its upper extremity it passes 

 insensibly into the young fruit, which in fact is only a continuation 

 of it. It is in reality in this dilated part of the peduncle, which 

 we have called the receptacle of the flower, and which is situated 

 below and around the disk of which we have just spoken, that the 

 principal increase takes place, at least in the greater number of 

 pears. It is then the peduncle itself which is here transformed 

 into the fruit, if we mean by this word the succulent and esculent 

 tissue, absolutely as in Anacardium and Hovenia. If any doubt 

 could remain about the matter, it would be removed by the exa- 

 mination of those abnormal fruits, like that represented by M. 

 Naudin in his " Note on the Structure of the Flower in Cucurbi- 

 tacese," which are real pears formed entirely at the expense of the 

 peduncle, since having neither heart nor carpels, nor vestige of 

 calyx leaves, they have never been terminated by a flower. 



" If I have made myself understood, we shall see that the struc- 

 ture of the ovary in the pear differs in nothing from that of the 

 ovary in other vegetables, and that it is altogether conformable 

 to the general plan of organization explained by our illustrious 

 masters, E. Brown, DeCandolle, and Jussieu. It is not then 

 necessary to bring forward the axis, to which appeal is made at 

 the present day so often and so willingly, when it is required to 

 explain the structure of flowers or fruit. I go further, and if I 

 am not deceived, it is not impossible to refer to the common plan 

 of organization, ovaries with a central free placenta, of which the 

 differences from ordinary ovaries would in this case be more ap- 

 parent than real. A strong presumption in favour of this mode of 

 view, if not an absolute proof, is afforded by the very homogeneous 

 family of Melastomacese, in which the most opposite modes of 

 placentation are found*. Thus, for example, in the suborder of 

 Melastomese, properly so called, whose ovary has from two to 

 twenty cells, the placentae are axillary, that is to say, altogether 

 adherent to the central columella which results from the line of 

 junction of the carpellary leaves ; in Ewyckia, where there are 

 four cells in the ovary, they are, on the contrary, parietal, or if the 

 phrase is preferred, fixed on the middle of the cells. Between 



* See ' Rumphia,' in which I published in 1834 the analysis of this family. 



