432 ANNUAL REPORT SMITHSONIAN INSTITUTION, 1961 



and the relative humidity is high. It is usually above 90 percent in 

 rain forests where fog is frequent at dawn. Much of the w^ater flows 

 oS the land into the large rivers that are characteristic of rain-forest 

 regions. In doing so it leaches the land and the surface soil becomes 

 denuded of salts. Forest waters are for this reason among the softest 

 in the world ; they may have a salt content only two or three times that 

 of rainwater. 



More accurately, it may be said that the nature of the environment 

 is controlled by the proportion (l/g) of the gain of water to the en- 

 viromnent in rain to its loss, not only in evaporation from open water 

 surfaces but in transpiration from the vegetation and percolation into 

 the soil. Where the rain is less than would be lost in these ways if 

 the water were there, the environment will be a dry one and no water 

 will flow ofi^ the land; where the gain is more than the loss, the en- 

 vironment will be humid. In the rain forests the value of l/g is of the 

 order of 0.2, so that much water flows off the land and the environment 

 is humid. In deserts l/g may be as high as 200, and in tropical grass- 

 land and savannas it is probably often near unity. 



Many other conditions in tropical environments are controlled by 

 the value of l/g. Small diurnal and annual ranges of temperature 

 are characteristic of humid environments, that is to say of those with 

 a low l/g, not only in equatorial regions but generally in the Tropics ; 

 in deserts the annual range may be as high as 40° to 50° C. and the 

 diurnal range 25° or 30° C. Ultraviolet light is less in the more 

 humid environments, cloudiness is greater, and the hours of sunshine 

 less (5 to 6 hours a day in rain forests) . 



This account of tropical climates is very summary and incomplete; 

 the few data I have given are almost wholly confined to the two ex- 

 tremes of climate, the rainforests and the deserts. But I hope that 

 it will serve to bring out some of the biologically important differences 

 between tropical and temperate climates. The clearest of these are, 

 besides the obvious difference in temperature, the much smaller sea- 

 sonal differences in tropical and especially equatorial climates, and 

 the greater part that w\ater supply plays in controlling the environ- 

 mental conditions. It is in fact true that in many tropical environ- 

 ments the effective rhythmical change of climate is not that of the 

 seasons but that between rainstorms. I have myself found this to 

 be clearly true in a country, the Paraguayan Chaco, where rain fell 

 at intervals of about a fortnight. Pools and other small bodies of 

 water would fill when the rain fell and dry before the next rain. Much 

 of the smaller fauna of these pools — such, for instance, as the branchio- 

 pod Crustacea, e.g., Estheria — passed through their whole life history 

 in the few days that the pools were full, hatching at the time of rain 

 and laying eggs before the pool dried. 



