:HAr. II] PERIODIC PHENOMENA IN THE TROPICS 253 



ropical dry seasons in the savannah-forests of Venezuela. Most of the 



rees were Icallcss during my visit in March, 1883 ; not a trace of 



egetative activitj- was visible on them, and yet many of them, in particular 



pecies of Cassia and other Leguminosac, were completely covered with 



owcrs. Epii)hytic bromeliads and orchids were also in full blossom. 



)n the other hand, the vegetation on the ground was almost flowerless ; 



was howe\'cr chiefly composed of grasses and other herbaceous plants, 



le plastic substances of which accumulate chiefly in the foliage and 



jnsequently must be expended for the production of flowers during the 



getative season, that is to say, during the rainy season. In March their 



liage was as dry as straw. 



A show of flowers often even richer, particularly of perennial herbs, 

 :companies the first rain after the dry season. But this continually 

 Ecreases as the rainy season proceeds, especially as regards woody and 

 iberous plants, and sinks to a minimum at the close of the rainy season ; 

 hilst the growth of the foliage-shoots still continues for a long time, 

 le secondary growth in thickness attains its greatest intensity, and 

 isimilation as well as other nutritive processes are at a maximum. 

 The fruits of many woody plants that have blossomed in the dry season 

 pen during the ensuing rainy season ; others require a longer time, 

 ct the ripening season for most fruits, so far as it is confined to an}- 

 irticular season, appears to be chiefly the rainy season. Accordingly 

 lere is very little fruit during the dry season. 



The favourable influence of the dry season on flowering is not at all 

 irprising. On the contrary, it is more a matter for surprise that certain 

 lecies of plants should blossom at the height and towards the end of 

 le rainy season. Such species are however very much in the minority, 

 pecially when only woody plants are taken into consideration. It has 

 :en shown in a former chapter ' that poverty of water in the soil and in 

 e atmosphere favours the inception and growth of flowers. Blossoming 

 the dry season, or soon after it, is a phenomenon that is intelligible on 

 lysiological grounds. Why on the contrary it should in other cases, 

 re however, be induced b}' the great humidity of the rainy season, 

 ay perhaps be correlated in the case of woody plants with certain 

 laptations, for example with special pollinating agents and the like, 

 hat, howe\er, many herbaceous plants without persistent stores of reserve 

 aterial should blossom during the rain, is a necessary consequence of 

 e direct dependence of the flowers on the foliage. 



iv. SPECIAL CASES. 



In order to obtain reliable and sufficient results regarding the influence 

 the tropical seasons on the formation of flowers, I have collected from 



^ See p. 26. 



