1896. DISPERSAL OF SEEDS BY BIRDS. 195 



a number of species very different in the size of flower. F. fragrans, 

 with small flowers in corymbs, has small orange berries, dis- 

 persed by birds and bats. F. zeylanica has large flowers and 

 globose grey berries an inch and a half in diameter ; too large for 

 our small birds, and thus adapted rather for dispersal by civets. 

 F. imperialis has enormous flowers, followed by correspondingly 

 large fruit six inches long and conical in shape. It would be 

 impossible for any of our birds to carry them away whole, so some 

 modification becomes necessary. The fruit, which is of a grey colour 

 outside, splits vertically, and the segments enclose an orange-coloured 

 conspicuous pulp easily accessible to birds, which must swallow the 

 minute seeds with the pulp. Unlike the fruits of F. zeylanica this 

 fruit is not readily detachable, as it would not be to the advantage 

 of the plant were it to fall off easily ; whereas in F. zeylanica the 

 fruit readily falls off whole without ever splitting, and it being 

 intended to be taken whole by nocturnal animals, the pulp is not 

 coloured. 



One may compare the arrangement in F. imperialis with that 

 in Moniordica Charantia, a pumpkin half wild, and cultivated all over 

 the tropics. This pumpkin is, in many forms at least, too large for 

 birds to carry off whole. It therefore spHts, and the segments 

 recurve and expose the seeds lying in a sweet crimson aril. 



In this case, however, the fruit itself is of a brilliant orange, and the 

 colouring of the segments sets off the crimson seeds well. Its appear- 

 ance suggests that at one time the fruit was small and eaten whole ; 

 but as it increased in size and its disseminator was unable to carry 

 the whole fruit off, it was so far modified that it practically became 

 capsular with arillate seed so as to suit the exigencies of the case. 

 In very many capsular fruits the seeds are provided with an eatable 

 aril of very variable size, sometimes so small that one wonders that 

 birds or mammals should find it worth while to seek it. Dysoxylon 

 canliflontm (Meliacese) is an excellent example. The capsules are 

 about an inch long, ovoid, and of a bright orange-red. They are 

 borne on the trunk of the tree, and, when ripe and split, stand horizon- 

 tally, the apex facing the spectator. The seeds are black and from 

 two to four in number. On the apex of the inner face of each is a 

 small but conspicuous orange aril not a quarter of an inch long. 

 This is the only part eaten, and I can perceive no taste in it, yet it is 

 a most popular fruit, being eaten, not only by birds, but also by 

 squirrels and Tupaias as soon as it is ripe. D. acutangnlum has a 

 very large orange capsule, as big as a small pear, in each cell of which 

 is a single large black seed with a yellow aril, forming a contrast of 

 three colours. Another species of Dysoxylon {D. angustifolium) is 

 chiefly remarkable for its capsule being creamy white. It splits at 

 the apex and discloses three or four seeds covered with a thin crimson 

 aril, which is very sweet. 



The split carpels of the Wormias (Dilleniacese) are nearly as showy 



p 2 



