1849.] 



exercised hy Trees on Climate. 



423 



regulate and equalize the flow of the water. But if on the other 

 hand the whole annual quantity of the flowing water becomes less 

 as the clearing away of the forests extends then the efiect must 

 be attributed either to the rains becoming less abundant or to its 

 evaporation being promoted by the ground not being covered with 

 trees and being thus deprived of shelter against the sun's rays 

 and the wind. These two causes, which always produce a similar 

 effect, must generally be combined. Before, however, endeavour- 

 ing to estimate their influence, separately, it will be proper 

 satisfactorily to ascertain if it be irrefragably established that the 

 water courses of a country, where clearing has been extensively 

 accomplished, are truly diminished, and thus assure ourselves that 

 the mere appearance of the phenomenon be not taken for its 

 reality. And, after all, this is the important point of the inquiry ; 

 for let it be once established, that the clearing of a country 

 reduces the quantity of water in its running streams, the cause of 

 this diminution is of secondary importance. We must now, 

 therefore, inquire if we can find in nature a set of phenomena which 

 will act as a criterion in resolving this question. 



I regard lakes, whether met with in plains or in different stages 

 of mountain chains, as highly qualified to throw light upon this 

 discussion. These may be considered as natural gauges calculat- 

 ed to assist in valuing upon a great scale, the variations which 

 may take plac% in the quantities of water which fertilize a country. 

 If the volume of these waters undergo any variation, more or less, 

 it is manifest that this variation, whether of excess or diminution, 

 will be indicated by the mean level of a lake inasmuch as its mean 

 level varies at different times of the year, according as the season 

 is wet or dry. Thus the mean level of a lake will fall, if the 

 annual quantity of running water in the streams of the district 

 diminishes ; on the contrary, it will rise, if these streams become 

 more copious ; and, finally, the level will remain stationary if the 

 volume of water which runs into the lake experiences no change. 

 In the following discussion, I have preferred the use of those ob- 

 servations only which relate to lakes which have no outlet ; and 

 the reason will be apparent as we wish to ascertain changes, which 

 may not be very considerable. However I do not disregard such 

 lakes as have an issue for their waters, for I am certain that the 



