2o6 Oranges and Lemons of India. 



44, fig. y) ; and the still more striking Dumrz form 

 of the " Khatta" (pi. 27, fig. 6). 



There is a point upon which I have not yet hazarded 

 any speculation : that is How could the oil cells of the 

 Citrus first have come into existence ? Were they 

 oil cells from the beginning of time, or were they 

 transformations of something else ? Taking into con- 

 sideration that these oil cells now exist all over the 

 bark ; that the axillae, or angles, of the leaf crenations 

 contain each an essential oil cell ; that in the Bryo- 

 phyllum calycinum buds are produced in similar 

 positions, and that in homologous points of the pulp 

 carpels seed-buds are developed, the suspicion might 

 be raised in one's mind that after all these oil cells 

 so constant in all parts of the Citrus may not impos- 

 sibly be connected with "spore" sacks, or similar 

 organs, in some ancestral cryptogamic form.* 



The same powers which were equal to changing the 

 five-toed animal into our one-toed horse, the lizard into 

 the bird of to-day, the lower animal into man, would, I 

 fancy, be also equal to bring about the changes I have 

 tentatively endeavoured to sketch in the foregoing 

 pages. 



In discussing the possible morphology of the citrus 

 fruit, a few words on its leaf may not be out of place. 

 The leaf is an expansion of the bark, specialised for 

 certain ends. It consists of two parts, the petiole (or 

 leaf stalk), and the leaflet (or blade). These are united 

 by a joint, like those of compound leaves, suggesting 

 the idea that the ordinary citrus leaf was at some time 

 a compound leaf. In fact, in the Citrus trifoliata, pi. 

 254, fig.g", we have a compound trifoliate leaf. In the 

 allied genus, CEgle marmelos, pi. 242, fig. f t we have 



* Vide note at the end of this chapter. 



