258 On the Separation of the Pomegranate 



[Oct. 



celled, with the placentas of the upper cells reaching from the parietes 

 to the centre, while those of the lower division proceed irregularly from 

 the bottom of the fruit. He does not, however, assign this peculiar struc- 

 ture as his principal reason for viewing the order as distinct from Myr- 

 taceee, but has recourse to others, in my estimation, of minor import- 

 ance. 



Lindley conceives that there are two rows of carpels, three or four 

 of which surround the axis at the bottom, while the remainder surround 

 these, and, occupying the upper part of the fruit, adhere to that part of 

 the tube of the calyx. The p acentas of these upper carpels, he con- 

 ceives, contract an irregular kind of adhesion with the back and front of 

 their cells. The meaning of this is far from being clear to me, but if 

 it means that he considers the plat entas of the upper as well as the 

 lower row to proceed from the axis towards the circumference, to which 

 last they contract accidental adhesions, then he takes an erroneous view, 

 and if the examples quoted in illustration support this view, they are 

 Dot in point as regards the structure of Punica. 



Mr. Arnott, like Lindley, vi^>ws the fruit as consisting of two rows of 

 carpels, an outer and inner, the former of which he thinks may be adven- 

 titious. To understand his theory, we must first suppose the tube of the 

 calyx spread out as a flat surface and covered with two circles of carpels, 

 the inner next the axis, and the other occupying a larger circle beyond, 

 that the margin of the cJyx then contracts so as to turn the outer series 

 over the inner. According to this supposition, the attachment or base 

 of the placentas of the outer series should be in the circumference, and 

 the apex in the centre, while that of the inner should be in the opposite 

 direction, that is., have the base in the centre and the apex towards the 

 circumference ; an explanation which is in accordance with what we 

 find, except in so far as it does not account for the horizontal partition 

 between the two series : nor can I exactly understand on what ground 

 we are warranted in assuming that the outer series is adventitious and 

 the result of cultivation, as it has every where been found so constant in 

 all circumstances. But be that as it may, this theory certainly accounts 

 for the crossing of the placentas in the two rows which we so invariably 

 find, whether correctly or not, cannot be determined, until we get fruit 

 with a single row of carpels, which has not yet been found. 



These explanations, which I venture to propose, of rather obscure des- 

 criptions, did not occur to myself until after I had formed a new theory 

 of my own, the result of a very careful examination of the ovary in all 

 stages from the earliest, up to the period of impregnation. At these 

 early stages, when the whole flower had not yet attained half an inch in 



