GOLD COAST REPORT ON FORESTS. 31 



been borne under the heading of cost of fuel, and the same 

 felling operations will then serve two purposes instead of only 

 one. Such planting operations should, however, if they are 

 expected to help towards the protection of the soil against ex- 

 posure, be undertaken with species with the evergreen habit and 

 not alone with those that drop their leaves in the hottest and 

 driest season of the year, as is unfortunately the case with the 

 i'ara rubber plant (I [even- brasiliensis) , one of the quickest 

 growing, hardiest, and best rubber producers known. This 

 species has the deciduous habit most pronounced when growing 

 in localities having a well - marked and prolonged dry se;is<m. 

 Nevertheless if grown in mixtures with cocoa (cocoa as the 

 ' underwood " and Para as the " overwood ") a sufficient protec- 

 tion should be afforded the soil by the dense shade cast by the 

 latter. The Para rubber trees are only bare of leaves for a short 

 period each year. 



The land at the disposal of the mines is, generally speaking, 

 sufficiently extensive to permit of a long felling rotation, and in 

 consequence the secondary growth has time to reach good dimen- 

 sions before it is again removed. This is not the case with the 

 farming rotations usually adopted by the natives, who return 

 to the same area in periods varying from three to nine years, and 

 the secondary growth has therefore insufficient time for attain- 

 ing a large size. Hence, looked at from the worst point of view, 

 the dangers following the cutting of the forests near the mines 

 for firewood are nothing like as great as in the case of the much 

 more extensive areas utilised by the native farmers. Their 

 operations, for the reason indicated above, are much more 

 likely to, and actually do, adversely affect the climate and tha 

 soil. If by the discovery of new gold fields the mines ever 

 become so numerous as to involve extensive and continuous tracts 

 of country, then the consequences of felling in a haphazard 

 manner every year siTch large blocks of forest would be more 

 serious, and the fellings would have to be regulated by Govern- 

 ment. 



After leaving Prestea we proceeded ??' Bromase to Ahirisu, 

 on the Mansi stream, a feeder of the Ankobra River, thence to 

 the village of Xikwansia, on the latter, and from there to the 

 valley of the Sibiri stream, a feeder of the same river. Another 

 long day's march westwards brought us into the drainage area 

 of the large and important Tano River. Several days were 

 spent in systematically exploring and analysing the forest vege- 

 tation of the An'totra River, and as it is more convenient to 

 describe the forests in groups according to the drainage systems 

 occupied by them, I will begin by giving a short sketch of those 

 found in the Ankobra basin. 



Forest Ver/efation of tlie Central Portion* of tlic Anl-nlwt 



Drainage. 



The character of the country and the type of forest vegetation 

 clothing that portion of the Ankobra drainage situated between 

 Tarkwa and Prestea has already been described. The remain- 

 ing tracts examined lie to the north of the latter settlement, and 



