10 TROPICAL NATURE, AND OTHER ESSAYS. 



during tlie years 1853 — 1862 shows a mean humidity 

 in January of 90, while the highest monthly mean for 

 the four years at Batavia was 88 ; and while the lowest 

 of the monthly means at Clifton was 79 '1, the lowest at 

 Batavia was 78*9. These figures however represent an 

 immense difference in the quantity of vapour in every 

 cubic foot of air. In January at Clifton, with a tem- 

 perature of 35° to 40° Fahr., there would be only about 

 4 to 4^ grains of vapour per cubic foot of air, while at 

 Batavia, with a temperature from 80° to 90° Fahr., there 

 would be about 20 grains in the same quantity of air. 

 The most important fact however is, that the capacity 

 of air for holding vapour in suspension increases more 

 ra]3idly than temperature increases, so that a fall of ten 

 degrees at 50° Fahr. will lead to the condensation of 

 about 1\ grains of vapour, while a similar fall at 90"" 

 Fahr. will set free 6| grains. We can thus understand 

 how it is that the very moderate fall of the thermometer 

 during a tropical night causes heavier dews and a 

 greater amount of sensible moisture than are ever ex- 

 perienced during much greater variations of temperature 

 in the temperate zone. It is this large quantity of 

 vapour in the equatorial atmosphere that keeps up a 

 genial warmth throughout the night by preventing the 

 radiation into space of the heat absorbed by the surface 

 soil during the day. That this is really the case is 

 strikingly proved by what occurs in the plains of 

 Northern India, where the daily maximum of heat is far 

 beyond anything experienced near the equator, yet, owing 

 to the extreme dryness of the atmosphere, the clear nights 

 are very cold, radiation being sometimes so rapid that 

 w^ater placed in shallow pans becomes frozen over. 



