42 Oranges and Lemons of India. 



the pummelo proper or shaddock of pi. 92, and in Pondi- 

 cherry, I found a still smaller pummelo, with a girth of 

 only twelve inches, that is, half an inch more than the 

 yellow orange of Kandy. Had there been evolved a 

 sweet and sub-acid amilbed, the further elephantisation 

 of this citrus in India would have become easy. Of the 

 amilbeds figured, one had eight pulp carpels, and nine 

 others from eleven to sixteen carpels, while of the pum- 

 melos proper, one had thirteen carpels, and six others 

 from fourteen to twenty-one, so that the hugeness of 

 the amilbeds and pummelos may be due, not only to 

 excessive growth of all parts, but also to a sort of 

 doubling, or proliferation of the carpels. In that case, 

 the outer skin would have either to burst, and so lose 

 one of its great factors in the struggle for existence, as 

 I have attempted to show in the chapter on Morpho- 

 logy, or it had to expand and make room for its mul- 

 tiplied contents. The normal number of carpels- 

 five I have only found in the diminutive qumquat ; 

 in all other citrus there has been a multiplication of 

 pulp carpels, until we reach the largest number in the 

 pummelos. 



In accounting for the size of certain citrus, we 

 should not lose sight of the possibility of their huge- 

 ness having been originated by a. fusion of two ovaries 

 In pis. 223 and 224 are figures of double lemons, 

 which I got from the Etawah jail garden, and from 

 the English shops ; one had no seeds ; but in the 

 English shops I came across two others, both of 

 which had seeds. It is therefore not unlikely that by 

 sowing seeds of double citrus, and by selection, 

 eventually a large and perfectly round fruit may be 

 obtained, which may have ceased to show any trace of 

 fusion of the two ovaries, except in the increased 

 number of carpels, and size of the fruit. Pis. 72 and 



