5o8 STUDIES IN SEEDS AND FRUITS 



be a species of Andira. A tree named the Hog-doctor, perhaps the 

 terebinthaceous Rhus Metop'ium^ supplies yam-sticks for trailing yams, 

 the sticks soon rooting and producing leaves. 



Several other Jamaican plants exhibit the same quality. Thus the 

 malvaceous tree, Paritium elatum^ is said to grow from branches stuck 

 in the ground, whilst the Dagger-plant [Yucca aloifolid) readily strikes 

 into the ground again when it is cut down. This not infrequently 

 occurs with some of the trees above mentioned. Thus on one 

 occasion I noticed that a tree of Cedrela odorata^ which had been lying 

 on the ground for some time after being cut down, was sendmg up 

 shoots. 



When in Fiji I found that the natives used the branches of 

 Barringtonia racemosa in constructing live-fences. When stuck in wet 

 soil they soon develop roots and foliage. Other trees are credited by 

 the natives with the same capacity, such as Thespesia populnea and 

 Hibiscus tiliaceus. 



NOTE 32 (p. 12). 



Twist Coco-nuts. 



This name is applied by the Tobago natives to dry nuts where the 

 kernel is loose and shakes when the fruit is handled. This may be 

 normal in many palms, as it is in Acrocomia of the same tribe ; but with 

 the coco-nut it is the exception. I regard the peculiarity as exhibiting 

 the tendency of the seed to enter upon a rest-period, the persistent 

 vital connection between the seed and the pericarp being a requisite 

 condition for true vivipary, as in Rhizophora. 



In such fruits, when broken open, the loose kernel comes out readily, 

 looking in its brown skin much like a huge Brazil-nut. When the 

 kernel is cut across, it proves to be without water in its central cavity, 

 is perhaps thicker than usual, though this is not always the case, and 

 contains a sound normal embryo. Mr H. Matthes, a planter of 

 " Bacolet," Tobago, kindly gave me the following particulars. 



Taken at random, out of a heap of coco-nuts about i or 2 per 

 cent, would display this character. Out of 1000 trees in one locality 

 10 produced these fruits. Generally a tree produces only twist-nuts, 

 with occasionally a few normal fruits. Of these ten trees all but one 

 stood in imperfectly drained situations, where water was apt to accumu- 

 late in the rainy season ; and to this cause he is inclined to attribute 

 their behaviour in this respect. This results in the retention of the 

 fruit on the palm, such nuts seldom dropping from the tree, but 

 undergoing all their shrinkage attached to the parent. When compared 



