THE WINDS. 277 



aiTested, drawn in, heated, and caused to ascend ; and so, the 

 northeast trade-winds are first weakened, then " killed," and after- 

 ward drawn into the vortex of ascending air over the burning 

 sands of the deserts ; on the other hand, the southeast trade- 

 wind, failing, when it arrives at the place where the equatorial 

 Doldrums were wont to be, to meet with them or any opposing 

 force from the northeast trades, are drawn over into the northern 

 hemisphere. Going now from the equator toward the poles, their 

 tendency is (§ 126) to obey the forces of diurnal rotation, as well 

 as those of the indraught for the heated plains, and thus the south- 

 east trades become southwest monsoons. In this view, the "equa- 

 torial Doldrums" of the Indian Ocean are transferred, as it were, 

 during the southwest monsoons, to the deserts of Central Asia. 



797. It may be asked by some saying, Since we can not always 

 tally the air, how do we know that these southwest monsoons are 

 the southeast trades of the Indian Ocean ? The reply is, We infer 

 that they are, because in co-ordinating for the Pilot Chart of that 

 sea we \\2iYQ foicnd (§ 795) no belt of ccthns hetvjeen the soiUheast 

 trades and the southwest monsoons^ but a gradual change, so to 

 speak, of the one wind into the other. Thus, confining ourselves 

 to August — one of the southwest-monsoon months — and to the 

 strip of ocean between 85° and 90° east, the investigation gives 

 as follows for calms and winds in the field between : 



10° S. and 5° S. 133 observations. calms. Wind S.E. 



5° S. and 0° 102 " 3 " " S. 



0° and 5° N. 99 " 3 " " S.W. 



5° N. and 10° N. 77 " " " S.W. 



798. These monsoons do not, as we are generally taught to 

 suppose, commence or end at the same time all over the Indian 

 Ocean. 



799. The Pilot Charts (Plate V.) have brought this fact out in 

 very bold relief. Take, as an illustration, the strip of ocean be- 

 tween the meridians of 85° and 90° east, south of Calcutta, and 

 as far as the equator. Let us divide it into "fields" (Plate V.), 

 by drawing across it lines to represent the parallels of 5°, 10°, 

 15°, and 20° north. 



800. In the first field below Calcutta, ^. e.^ between the land 



