DISTKIBnTION OF THE RAINFALL. 283 



Such a distribution of the rainfall is evidently attributable 

 primarily to geographical situation, and thus may it be with the 

 rainfall generally. But the observations of M. Cezanne and of M. 

 Eaulin seem to show that irrespective of forests the distribution of 

 rainfall dependent on geographical position may be determined by 

 the contour of a country. And when, in view of this, we look to the 

 correspondence between the distribution of the rainfall which has 

 been adverted to, we see no necessity for concluding that it has been 

 the existence of the forests which has determined the corresponding 

 distribution of the rainfall in the places cited, as that seems to 

 exercise a much more powerful influence than this can. 



Sect. III. — On ike Distribution of Forests Affected hy the Distrihv.tion 

 of the Rainfall. 



The general accordance between the distribution of forests and the 

 distribution of the rainfall on the earth's surface, detailed in the first 

 section of this chapter, naturally suggests, as has been stated, the 

 enquiry, Can they be consequences of some common cause? or is 

 this general accordance only an incidental coincidence ? or are they 

 connected together as being one of them the effect of the other as 

 the cause or the occasion of its appearance ? And if so which of them, 

 the distribution of the rainfall or the distribution of the forests, which 

 is the cause and which is the effect ] To what extent is the one the 

 cause or occasion of the appearance of the other ? and may not that 

 which may appear to be the effect have a reflex influence upon that 

 which appears to be the cause or occasion of the other ? 



Such is the question raised, and I have adduced evidence to show 

 that the distribution of the rainfall dependent on geographical 

 position may be determined in a great measure by the contour of a 

 country, the elevation and the configurations of its mountains, its 

 valleys and its plains, irrespective of whether it be clothed with 

 woods or altogether devoid of these; and in a preceding section 

 I have stated my opinion to be that the distribution of the rainfall 

 may primarily have determined the distribution of the forests, but 

 that the forests once estabhshed may have exercised an important 

 influence on the distribution of the rainfall. 



It is stated by Mr Marsh : " There is good reason to believe that 

 the surface of the habitable earth, in all the climates and regions 

 which have been the abodes of dense and civilized populations, was, 



