2 FORESTS, WOODS, AND TREES 
and stream-flow, and on the erosion of the soil on slopes of 
hills and mountains. The positive results of these observa- 
tions, which are detailed below under separate headings, 
may be considered to hold good with regard to the British 
Isles and temperate regions generally. It is only fair to 
state that the beneficial action of the forest in increasing 
the rainfall, in diminishing the run-off water, and in pre- 
venting the erosion of the soil is not universally admitted. 
The main arguments against the ordinary view were 
published by H. M. Chittenden (1) in 1908. The attention 
of foresters and engineers may also be directed to the im- 
portant monograph, entitled Boschi e Acque, which was 
published at Rome in 1916 by M. Giandotti (2), Director 
of the Hydrographic Office of the river Po. This is a 
complete study of the whole question of the relation of 
forests to rainfall and stream-flow. 
The investigations carried out in India on the influence 
of forests on atmospheric and soil moisture, which have 
been the subject of a recent official report (Indian Forest 
Bulletin, No. 33, 1916), throw no new light on the problem 
as regards tropical regions. The general conclusions arrived 
at in India were: “The influence of forests on rainfall is 
probably small, but the denudation of the soil, owing to 
the destruction of forests, may be regarded as an established 
fact in India.” Dr. Gilbert Walker, in an appendix to 
this bulletin, points out the difficulty of such investigations, 
one cause of error being the tendency of the annual rainfall 
to run in spells of excessive and deficient years, so that the 
exact influence of forest growth or of forest destruction is 
not readily arrived at. Dr. Hugh R. Mill, in Nature, 
2nd August 1917, p. 446, advocates a study of the relation 
of the isohyetal (3) lines to the configuration of the land 
on wooded and treeless districts of similar character; and 
instances from the report of the rainfall in the Geological 
Survey’s Water Supply Memoirs on Hampshire that the 
district of the New Forest shows a considerably higher general 
rainfall than its elevation above sea-level appears to suggest. 
It will be convenient to state now under four distinct 
