256 DISTRIBUTION OF RAINFALL AND OF FORESTS. 



Sect. I. — On the Measure of Correspondence between the Distribution 

 of the Rainfall and of Forests. 



I have said it appears to me that observations made show in many 

 cases a general correspondence in the distribution of the rainfall and 

 of forests. It may occur to many on the mention of this that while 

 there are found forests covering extensive plains, their special habitat 

 appears to be the mountain side — and that the rainfall in the vicinity 

 of a mountain range is in general greater than it is on the horizontal 

 area of an extensive plain. 



But the correspondence is more marked than this alone would inti- 

 mate. It happens to be the case that, in some cases, the observations 

 have been given to the world by observers having decided views in 

 regard to the connection of cause and effect, and in illustration of 

 their views ; but at this stage they are adduced only as illustrative 

 of the measure of correspondence which has been observed. 



The first observations I adduce are some brought before the Meet- 

 ing of the British Association for the Promotion of Science, held 

 at Brighton in 1867, by Dr Brandis, Superintendant-General of Forests 

 in India, in a paper On the Geographical Distribution of Forests in 

 India. Dv Brandis considering India as divided, by observation of 

 the rainfall, into arid, dry, and wet districts, stated that in the west 

 corner of India was what might be called the arid tract, extending 

 from the coast of Cutch and from Scinde in the south to the Salt 

 range in the north, and from the hills of Beloochistan in the west 

 to the Aravalli range in the east. The average rainfall in this 

 district was less than fifteen inches. Throughout this arid part of 

 India the spontaneous arborescent vegetation was extremely scanty, 

 although a thin sprinkling of low thorny scrub on the hills offered 

 ample and interesting employment to the botanist. In this region 

 the work of the forester was limited to those tracts which stretched 

 along the Indus and its principal tributaries, watered by the annual 

 overflow of the river during summei-, or which could be otherwise 

 irrigated. Thus in Scinde there were on both sides of the Indus 

 river 352,000 acres of Government forest maintained solely by the 

 overflow of the river and by percolation. The result of the deflec- 

 tion of a river from its course was that the forest near the old bed 

 frequently perished. 



Outside this arid tract there were two belts, with an annual 

 rainfall of between fifteen and thirty inches, which might be called 



