2OO Oranges and Lemons of India. 



but still a very primitive citron. This stage of fruit 

 evolution having been gained, it is not perhaps diffi- 

 cult to pursue the investigation to its possible comple- 

 tion, according to my views. The inner whorl of rind 

 carpels would naturally contain a^ hollow space, as in 

 these digitate citrons (see c (A), pi. 139). The oil-cells 

 of the outer whorl could not project beyond the rind 

 of the inner whorl ; but those of the latter could be 

 prolonged and projected into the hollow interior. 

 This circumstance, I take it, was favourable for the 

 development of the elongated juice vesicles from the 

 short oil-cells of the inner whorl, that is for their be- 

 coming larger and more differentiated. The oil-cells 

 of the outer whorl, in the perfect orange, are often 

 large, orange coloured, and project inwards like bal- 

 loons (see pi. 122, fig. /&). There is no trace of the 

 regular oil cells in the pulp carpels of the perfect 

 citrus, and therefore, supposing the latter to be the 

 real carpels, the lobes or vesicles, which project 

 inwards from them, as I have said, can logically 

 hardly be others than transformed oil-cells. If we 

 could forget the term essential oil-cell of the rind 

 and leaf, and replace it by pulp-vesicle-germ, we 

 might perhaps have little difficulty in believing all 

 three to be homologues. In order to complete the 

 history of the citrus fruit, according to my views, I 

 have something further to suggest. 



I believe that before the orange or lemon could 

 have been perfected, as we see it in our orchards, the 

 inner whorl of rind carpels must have developed some 

 sort of buds in the axillae of their crenations (or what 

 would correspond to them), which, in course of ages, 

 through the strictest selection, were perfected into the 

 present citrus-seed. Otherwise it is difficult to con- 

 ceive how variations of sufficient importance to cause 



