Chap. II] PERIODIC PHENOMENA IN THE TROPICS 251 



do Sul blossom at intervals of about thirteen years. Bambusa arundinacea, 

 on the west coast of Cisgangetic India, blossomed at intervals of thirty-two 

 years — 1804, 1836, 1868 l . According to Ridley, two species of Hopea, 

 H. intermedia and H. Mengarawan, and four species of Shorea, S. lepro- 

 sula, S. parvifolia, S. pauciflora and S. macroptera, blossom with great 

 regularity every sixth year. These cycles are said to coincide with very 

 dry years 2 . 



In most cases, during the greater part of the reproductive period, there is 

 a retardation or even a stoppage in the vegetative domain, and this may 

 extend to the whole crown, when its habit is to break simultaneously into 

 flower, or it may be confined to the larger or smaller branches, according 

 to their degree of individuality. The effect on the vegetative region is 

 frequently limited to the discontinuance of the formation of foliage-shoots ; 

 the vegetative buds rest. In many cases, the antagonism between the 

 vegetative and reproductive functions goes further. A tree or shrub pre- 

 paring to blossom throws off its foliage, chiefly hoivever from the flozvering 

 branches, whereas the purely vegetative ones usually retain their leaves. 



At what stage of the development of blossom this phenomenon happens I have 

 unfortunately omitted to determine, and the literature of the subject has nothing 

 to say about it. In this case, as in the shedding of the foliage of Urostigma gla- 

 bellum 3 , the effect possibly depends on the diversion of the transpiration-current 

 towards the flower-buds. The fresh foliage sometimes shoots out at the com- 

 mencement, sometimes at later stages of the formation of the fruit. 



Criiger had already observed in Trinidad that Erythrina blossomed 

 when bare of leaves and that twigs that remained flowerless retained their 

 foliage. I have frequently had an opportunity of seeing this statement 

 confirmed and have observed the same fact repeatedly in Schizolobium 

 giganteum in Java. I met with similar phenomena in the botanic garden 

 of Tjibodas, where I paid more attention to it, in Parasponia parviflora, 

 the richly flowering twigs of which threw off the greater part of their 

 foliage, whilst those that bore only a few flowers retained considerably 

 more leaves. I also noted it in an Ardisia and in Juannuloa aurantiaca, 

 where the blossoming twigs were altogether or nearly leafless, whilst a 

 reduction in the foliage was not exhibited on the purely vegetative twigs. 



I found also in many otherwise deciduous trees that the flowering twigs 

 acquired their leaves later than did the purely vegetative ones. Thus, on 

 November 21, 1889, in the botanic garden at Buitenzorg, I saw two trees 

 of Firmiana colorata with young leaves and flowers. The flowers were 

 abundant on one of the trees and scanty on the other, but in both cases 

 were confined to separate systems of branches. On the flowering boughs 



1 Brandis, II, p. 90. 2 Ibid. p. 20. 



3 See p. 245. 



