252 ZONES AND REGIONS [Pt. Ill, Sect. 



the leaves were still small and pale, but on the sterile ones they we: 

 already large and bright green. On December 13, after the flowerini 

 period, the tree that had flowered abundantly could be easily distinguishe 

 from the other by its less developed foliage. In Meliosma lanceolata, 

 the same garden, on November 31, I saw young foliage only on steri 

 boughs, whilst the twigs provided with infructescences or young inflo 

 escences still showed no signs of foliage. 



Matiy trees in their youth, so long as they do not produce flozvcrs, ai 

 evergreen, zvhilst later on they shed their leaves before the impcndii 

 blossoming period. This is the case, for instance, with Schizolobium gigai 

 teum, at least in Java. 



It is evident from the foregoing remarks that, like leaf-formation ar 

 leaf-fall, the development of flowers depends on a periodical!}- recurrir 

 internal condition of the plant. Shoots that flower continuously no mol 

 exist than do those that are continuously forming foliage. /;/ the rcpr 

 ductive doviaiji, then, there occurs a rhythmic alternation of rest and activi 

 depending on internal causes. 



iii. PERIODICALLY DRY DISTRICTS. 



The rhythm that is witnessed in leaf-formation is observable also 

 the flou^er. The production of floivers exhibits a correlation with t 

 seasons of the year, ivhenever the seasons display sharply defined different 

 In the reproductive domain this dependence is likewise a seconda 

 feature — an adaptation to external factors on the part of physiological 

 necessary processes. In the tropics an influence associated with variatio 

 in temperature is exhibited only in border-districts, and consequently ne 

 not be considered here. Over the greater part of the torrid zone, t. 

 difference in the seasons, as far as these concern plant-life, is expressJ 

 only in the atmospheric precipitations, and in particular in the rainfall ai 

 the atmospheric humidity. 



Tlie blossoming of zvoody and tuberous plants everyzvherc zvithin the trap 

 IS vu^st abundant during the dry season, or immediately after it ; and the 

 are precisely plants in which the production of flowers is not direct 

 dependent on the foliage. We frequently find it stated in the accour: 

 of travellers, as a remarkable phenomenon, that manj- trees blossc 

 precisely in the dry season. Belt makes this statement concerni:' 

 Nicaragua, Crliger concerning Trinidad, Schweinfurth concerning Nub 

 and Kurz saj's of the deciduous forests in Pegu, that most of the tre 

 blossom during the hot dry season, that a number of plants with rhizon^ 

 and tubers — for instance, Scitamineae, Amaryllidaceae, Orchidaceae, Och 

 suffruticosa — at the same time burst into blossom, and finally that t 

 leafless dried branches of the trees are covered with flowering orchids. 



I became personally acquainted with the abundance of blossom duri 



