Appendix. 349 



the juice vesicles arose also from the sides of the carpel (see 

 pi. 49, figs b and //), as ought to be the case, supposing they 

 were homologous to the oil-cells of the leaves. In the latter, 

 the oil-cells are disseminated all over the leaf, and not only 

 near the midrib. It is nevertheless a fact that generally the 

 juice-vesicles arise from that part of the carpel which goes 

 to form the circumference of the pulp-ball. This circum- 

 stance I shall endeavour to explain in the following way : 



The centre of the Hesperidium, or Citrus fruit is fixed, 

 while its circumference moves outwards as the fruit grows 

 and expands. We shall suppose that in the ovary, or when 

 the hesperidium is quite young, the oil-cells of the carpel 

 are on a level with its edges, or nearly so. It will be evident 

 that, the edges of the carpel, forming the centre of the fruit, 

 being fixed, some part must give way as the blade of the 

 carpel is pushed outwards by the growth of the juice vesicles. 

 Therefore those portions of the carpels, which have stretched 

 gradually to adapt themselves to the filling of the vesicles 

 with juice, will be those which now form the sides of the 

 carpel, and which are usually free of vesicles. (See pi. 106, 



fig. *) 



This giving-way sometimes goes beyond mere stretching^ 

 and the parts are torn asunder by the bursting power of the 

 expanding vesicles. This occurs generally in the suntara 

 and pummelo groups, which see.* In these cases, not only 

 the whole carpel is separated from the central column, but 

 often the edges of the carpel itself are torn asunder. 



In pi. 61, fig. b, the central column remains isolated at A, 

 while all the carpels have been torn away from it, and only 

 keep up a communication by means of the radiating fibres. 

 The carpels themselves are all closed. 



PI. 63, fig. b, shows the central column dragged to the one 

 side by a closed carpel, while almost all the other carpels 

 have not only receded from the centre, leaving there a hollow 

 space, but are burst open with the points of their juice- 

 vesicles projecting into the hollow centre. 



PI. 8 1 shows this bursting force in a more exaggerated 

 manner, tearing the central column into several parts, A, A, A, 



* One of the distinctive characters of the Portugal orange, and allied 

 forms, is that, at maturity, the centre is solid. 



