CilAr. II] PERIODIC PHENOMENA IN THE TROPICS 249 



The separation in point of time of the blossomins^ and of the purely 

 i-egetative condition is due to the fact that botli arc associated with 

 :liffcicnt seasons of the year. In the tropics also, an influence on the part 

 of the season occurs wherever a sharply marked change of climate prevails, 

 but this influence is usually weaker than in zones with a cold winter. The 

 uimber of species in blossom throughout the whole year is greater, and the 

 period during which blossoming individuals of one species are found is, in 

 general, longer in the tropics than in places where the seasons of the year 

 exhibit very great differences of temperature ; in fact, repeated blossoming 

 at short intervals, which in the temperate climate is almost always anoma- 

 ous, is a normal and regular phenomenon with many tropical plants. 

 Such distinctions are most striking when the woody plants of the temperate 

 and tropical zones are compared with one another. 



ii. CONSTANTLY HUMID DISTRICTS. 



The less marked tlie climatic differences of the seasons are, the less dependent 

 upon the season of the year is the blossoming period of tropical plants. 

 We find therefore the same relationships as in the vegetative system. 

 Species that blossom throughout the year are commonest in districts with 

 a nearly uniform climate. The remark frequently met with in tropical 

 Floras, ' blossoms throughout the year,' is however as a rule not to be 

 interpreted as meaning that one and the same plant is always in flower, 

 but merely that blossoming individuals are to be found at any time. 

 Among the species belonging to the category of plants that are constantly 

 in bloom, many occur the individuals of which bear blossoms only once 

 a year, or not even annually. This is especially the case with many trees 

 of moist virgin forests that are seldom found in blossom. On the other 

 hand, there are species, individuals of which possess a blossoming period 

 that is very long or recurs at short intervals, so that the number of their 

 blossoming plants at any time of the year exceeds that of those out of 

 blossom. This condition occurs in particular in open sunny situations and 

 in littoral forests. Species of Rhizophora and Avicennia, but especiallj- 

 Hibiscus tiliaceus, have remarkably long blossoming periods and are usually 

 found in bloom. I cannot state whether there are any long-lived plants, 

 individuals of which are uninterruptedly in bloom, since observations 

 extending over years would be required to determine it, and no one has 

 thought of doing this. I do not however consider such a condition 

 impossible, especially in the case of richly branched woody plants ; some 

 widel\- cultivated species of Hibiscus that seemed to me to be continuously 

 in blo.ssom. and Ricinus, should prove adapted for an easy solution of the 

 problem. But even then, there could not fail to be an alternation of 

 periods of rest and activity in blossoming. We have seen in the pheno- 



