1 882.] Forests— their I ■ . dfe and Rainfall. 2$ 



face. This will appear quite plain when it is remembered that 

 the extent of the leaf surface was six times as great as that of the 

 soil, and that the total diurnal evaporation was so nearly equal 

 from the two sources. These experiments were several times 

 repeated, and with about similar results. 



Now if it were known how many times greater the leaf surface 

 of a great forest than the land on which it was situated, it might 

 with ease be computed what is the relative evaporation from a forest 

 and an equal area of open country. From personal observation 

 and computation, we think it safe to assume that the leaf surface 

 of a wood is at least twelve times greater than the ground on 

 which it stands, so that at the above rate the transpiration from 

 the forest would still be nearly twice as great as the evaporation 

 from an equal area of free soil. It should be mentioned also that 

 the evaporation from the earth in this case was under the most 

 favorable circumstances, and the state of the ground as regards 

 moisture was very like that of the open earth directly after a 

 moderate rain. It was found by testing to be nearly equal to 

 that given off by a similar area of water. 1 It would appear cer- 

 tain, then, from these investigations, that more water is emitted to 

 the atmosphere from a forest than from an equal body of water, 

 and in this there is a confirmation of the experiments of Wil- 

 liams who computed that the evaporation from a wood was one- 

 third more than an equal space covered with water. 2 It is well 

 known that at times, during the warm season more particularly, 

 we have no rain for several weeks, so that the mean general sur- 

 face evaporation is probably not by any means as great as would be 

 indicated by these figures— for it was found that by allowing the 

 soil in the pot to become even moderately dry, the amount evap- 

 orated would fall far short of what it was when keeping the soil 

 well watered. On the other hand we have good reasons for be- 

 lieving that the true rate at which forests give out aqueous vapor 

 is, at all events, not over estimated in these researches. In the 

 first place the trees are at all times supplied with a more abundant 

 supply of moisture for transpiration — owing partly to power 

 which the roots have to attract moisture from every direction ; 

 partly to the retention of the rainfall in their network to be in due 



