APPENDIX 497 



seeds (78 grains), of which 3 had aborted, the proportional 

 weight of the pericarp was 97-8 per cent. The number 

 of seeds varies greatly and affects the relation. Thus, if the 

 Mandarin Orange referred to in the table had had a seed in 

 each of its ten segments, the proportional weight of the pericarp 

 would have been reduced from 99 to 98 per cent. 



(17) Cocas, pp. 325, 327. (See under Palms, p. 499.) The data here 



given for Cocos nucifera refer to the green coco-nut and to the 

 completely air-dried fruit without water, and can only be used 

 as indicated under Palms in this note. 



(18) Elais guineemis, p. 327. — See under Palms (pp. 499, 500). 



(19) Entada polystachya, p. 325. — See Note 26 on p. 494. 



(20) Erythrina corallodendron, p. 327. — ^The beading of the pods, which 



is dealt With on p. 360, has an important effect here. 



(21) Faha vulgaris (Broad Bean), p. 326. — The pericarp-proportion of 



a two-seeded dry legume, weighing 55 grains, was 26 per cent. 



(22) Gossypium (Cotton plants), p. 327. 



Proportion of Parts in the Dry Dehiscing Fruit. 



The wool forms 28 per cent, of the resting seed in G. barbadense, and 

 43 per cent in G. hirsutum. 



The influence of the hairs on the proportional weight of the 

 seed-coats is shown in the table on p. 194. In all cases 

 the fruits designated Gossypium hirsutum in this work are 

 from the same plant, wheth^ or not marked by (?) in 

 the text. 



(23) Hibiscus esculentus, p. 327. — The data given in the table refer to a 



dry fruit weighing 100 grains with 10 carpels and 94 seeds. 

 In a dry fruit weighing 62 grains with 7 carpels and 54 seeds 

 the pericarp- proportion was 36. 



(24) Hura crepitans, p. 325. — The Hura difficulty is discussed in 



Note 24 (p. 490). The drying of the entire fruit is discussed 

 on p. 289. 



32 



