THE FACTORS OF DISTRIBUTION 31 



received at the upper limits of our own atmosphere, 

 probably little more than half reach the surface of the 

 globe itself, most of those which are absorbed during their 

 passage through the atmosphere being taken up by the 

 dust particles upon which moisture condenses to form 

 cloud. The atmosphere is itself so nearly transparent 

 to the direct rays of the sun that, as is well known, the 

 sun may blister an alpine climber when the air around 

 him is below the freezing-point of water. Much of the 

 direct heat of the sun will be absorbed by the leaves of 

 plants upon which it falls, though not as large a pro- 

 portion as by the blackened bulb of a thermometer; 

 and it is well to bear in mind that many plants live in 

 the sun, while our records of temperature are mainly 

 taken from instruments kept in shade. The tempera- 

 ture of the ah* depends upon the amount of the heat 

 reflected from the earth's surface which is absorbed by 

 the moisture in the air. The amount of this reflection 

 differs greatly according as the surface is land or water. 

 Land reflects more of the heat that falls upon it, while 

 the high specific heat of water causes it to render much 

 of its supply of heat latent. The seas, or other bodies of 

 water, are thus more slowly heated than the land, and 

 retain their heat longer, giving it to the air more slowly, 

 and so lowering the general temperature of the air in 

 summer and raising it in winter, and producing in 

 their neighbourhood those equable or " insular " climates 

 with but little difference between summer and winter 

 temperatures, or " hiberno-aestival variation," which 

 contrast with the extreme or " continental " climates in 

 the interiors of continents. 



The first step towards a scientific geography of plants 

 was taken when Humboldt, in 1815, proposed to trace 

 over the earth's surface isotherms, or lines joining places 

 having the same temperatures. The extent to which 

 these lines fail to run parallel to the parallels of latitude 

 is a measure of the extent to which the distribution of 

 land and water modifies the latitudinal distribution of 

 heat. The isotherm of the mean annual temperature 

 of 82 F. corresponds roughly with the Equator ; but the 

 position of monthly isotherms varies with the seasons, 

 i.e. with the apparent northward or southward passage 

 of the sun, from being vertical over the Equator at the 

 Equinoxes (Spring and Autumn) to being so over either 



