86 Oranges and Lemons of India. 



In my opinion, with every generation that passes 

 without reversion, the power of reverting will become 

 weaker, until it is entirely extinguished. 



There is probably nothing in connection with the 

 citrus better established than that it has descended from 

 a progenitor with a compound leaf. The Citrits tri- 

 foliata plainly indicates this ; allied genera, the ^gie 

 marmelos, the Feronia elcphantum> and Limonia acidis- 

 sima and others point in the same direction. More- 

 over, the almost universal joint between the leaflet and 

 petiole of all the cultivated citrus (only absent in some 

 leaves of some citrons, and the two first leaves of young 

 citrus seedlings) prove it to have descended from an 

 ancestor with a compound leaf. Nevertheless, of some 

 hundreds of seeds of about fifty varieties of citrus that 

 I have sown, only three produced a few leaves with a 

 trifoliate character. These are shown in pi. 246, figs. 

 d, e,f*xAg. 



Is it, however, a fact that the petiole of the true lime 

 never reverts to some ancestral and larger form ? In 

 pi. 226, figs, d, e, f and h, I have given specimens of 

 true lime leaves, with large wings to their petioles ; 

 and in pi. 233, figs, a and b, I have given others taken 

 from seedling trees about five years old. 



I may as well note here that Loureiro's Citrus 

 limonum (" Flor. Cochin Chin. ") is given as a variety 

 of C. medica, and he refers the reader to Rumphius's 

 " Flora Amboyn.," vol. ii. tab. 29. Unfortunately this 

 plate gives two very different things, viz. : fig. a, which 

 I believe to be a lemon with unwinged petiole, and 

 probably descended from C. medica ; and fig. b, a true 

 lime, like the Indian kinds, and pictured with four 

 petals. Rumphius says that on rare occasions the a 

 form is produced, but he does not mention whether on 

 the lime tree proper, or on a totally different tree. 



