1 84 Oranges and Lemons of India. 



in fig. a, pi. 1 8 1. Then in fig. c, pi. 209, I have given 

 a portion of the surface of the saddphal, at A. The 

 largest cells are disposed at equal distances, shown by 

 the larger dots, the smaller oil cells occupying the 

 intermediate spaces as miliary eminences (see also 

 pi. 75, fig. a, A). In pi. 47, figs, c and e, are given 

 exact representations of sections of a citrus rind. The 

 larger depressions are the openings of the larger oil 

 cells, and the secondary depressions, the openings of 

 the second sized cells, while the third sized cells are 

 often so squeezed together as to be projected into little 

 convexities between the other openings, instead of 

 having each a depressed opening to itself. In small 

 specimens, the third sized cells are either mere 

 points, or not traceable. The only difference between 

 the margin of the rind, and the margin of the leaf, 

 is that in the former, the oil cells are much more 

 developed, and their essential oil differs from that of 

 the leaf. 



It would appear this is not the sole relationship be- 

 tween the leaf and the rind. In many cases the out- 

 line of some of the leaves corresponds closely to the 

 outline of a longitudinal section drawn through the 

 centre of the fruit ; and I believe that the mammilla of 

 the citron, close to the stalk, is only a representative 

 of the agglutinated winged '- petioles of the original 

 carpels, and that the opposite mammilla on the distal 

 end of the fruit may be an agglutination of similar 

 processes in the original carpels. I have come across 

 leaves of citrus, the outlines of which represented at 

 their tips the outline of the mammilla in the fruit.* 

 Therefore in cases where there is no distal mam- 

 milla, and the apex is depressed, it is not impossible 



* More likely the distal mammilla represents the agglutination of 

 the acuminate ends of the carpels. 



