May 26, 1887] 



NATURE 



^5 



UPPER WIND-CURRENTS NEAR THE EQUA- 

 TOR, AND THE DIFFUSION OF KRAKA TAb 

 DUST. 



'X*HE crude idea that the trade-winds on either side of 

 ^ the equator met in the doldrums, and that then the 

 air rising upwards flowed backwards as a return south- 

 west current in the northern hemisphere, and as a north- 

 west current in the southern hemisphere, has been modi- 

 fied by the more modern discoveries that the atmosphere 

 is not composed of horizontal layers of air moving in 

 different directions, but that, as a rule, there is a regular 

 continuous successive veering of the wind as we ascend. 

 It is not usual to find a southerly wind on the surface, 

 and for some height above, and then abruptly a westerly 

 current also of a certain thickness ; for cloud-observa- 

 tions show that over a surface south wind the upper 

 currents may be from south-by-west at quite a low level, 

 from south-south-west a little higher up, then suc- 

 cessively from south-west and west-south-west as we 

 ascend, and perhaps from west at the altitude of the 

 highest cirrus. We must, in fact, look upon the atmo- 

 sphere as circulating in the form of a continuous complex 

 screw. 



Innumerable observations show that, as a rule, there is 

 a very definite law of the vertical succession of the upper 

 currents. Stand with your back to the surface wind, and 

 the upper currents will come successively more and more 

 from your left hand the higher they are. The rule is 

 reversed in the southern hemisphere, for there the upper 

 currents come successively more and more from the 

 right. For instance, with a southerly wind in London, 

 the clouds will come more and more from the west the 

 higher they are, while in Australia they would come more 

 and more from the east. 



But during my two meteorological voyages round the 

 world I have discovered some very remarkable excep- 

 tions to this law in that interesting part of the world, 

 meteorologically, that lies between the equator and the 

 doldrums. These have so important a bearing, not only 

 on the whole problem of the circulation of the atmo- 

 sphere from the equator to the Pole, but also on the 

 remarkable diffusion of dust from the volcano of Kraka- 

 tab, that I propose to give a short account of these 

 researches in this article. 



In the Atlantic, the doldrums lie north of the equator 

 at all seasons of the year. Between 20° and 30° W. 

 longitude they range from about 3° N. ; in winter, to 

 11° N. in summer time. In the Gulf of Guinea, the 

 position of the doldrums cannot be accurately defined, 

 but they probably range from about 5° to 10^ N. of the 

 equator. There is, however, a striking difference in the 

 direction of the winds between the equator and the dol- 

 drums in the eastern and western portions of the Atlantic. 

 In the Gulf of Guinea, the south-east trade turns to south- 

 west and forms a south-west monsoon ; while west of 

 about 15° W. longitude the south-east trade remains from 

 nearly in the same direction as before it crossed the line. 



In the Indian Ocean, the position of the doldrums 

 varies enormously at different seasons. From about 

 November to March the doldrums are 5° to 10° S. of the 

 line, and the north-east monsoon draws into the well-known 

 north-west monsoon as it crosses the equator. From 

 April till September the doldrums must be somewhere to 

 the north of the line ; but, contrary to the opinion of 

 Dove and others, the Indian meteorologists now believe 

 that the south-west monsoon is not linked up regularly 

 with the south-east trade. As the evidence for this belief 

 is not yet published, we can say nothing about it, though 

 we shall have to refer to the point later on. 



The words rotation, circulation, and veering of wind 

 are unfortunately used so vaguely by different writers that 

 it may be well to define them more precisely before we 

 discuss the general circulation of the equator. 1 



" Rotation " should be confined exclusively to the 

 manner in which the surface wind blows round areas of 

 high or low pressure. For instance, we may say that the 

 wind rotates counter-clockwise round a cyclone in the 

 northern hemisphere. 



" Circulation " should be applied to the general move- 

 ment of a whole mass of air extending over a certain 

 breadth and height. For instance, the air in a cyclone 

 rotates round and in below, round and out at high levels ; 

 and the whole system makes up the general circulation of 

 a cyclone. Or, again, the whole motion of the atmosphere 

 into the doldrums, then upwards and more or less pole- 

 wards, should be called the general circulation in the 

 vicinity of the equator. 



" Sequence" of wind should be applied to the changes 

 in the direction of the wind which take place during any 

 interval of time as a cyclone passes over a station. This 

 veering or backing of the wind has unfortunately been 

 too often called the rotation of the wind. The confusion 

 here, as often, arises from sometimes talking of the 

 motions of a mass of air extending over a large area at a 

 given moment, and sometimes of the successive motions 

 of the air at a single station during a certain interval of 

 time. Some writers make matters worse by not only 

 drawing the plan of a cyclone on a chart, but also by 

 putting in small circles at different stations to show how 

 the wind would go round as the cyclone drifted past. The 

 confusion is direful. 



"Vertical succession" should be used to denote the 

 gradual successive directions of the upper currents, as 

 has been already explained. 



The two sections of weather across the Atlantic which 

 I made were taken, the one in July, between Rio Janeiro 

 and Teneriffe ; the other in December, between Tene- 

 riffe and Capetown. A short account of the observations 

 were published in Nature, vol. xxxiii. p. 294, so that the 

 general results only need be stated here. 



Between Rio and Teneriffe, while south of the line, the 

 low or middle clouds over the south-east trade invariably 

 came from some point to the right of the surface when 

 you stood with your back to the wind ; i.e. if the surface 

 wind was south-east the low clouds would drive from 

 about east-south-east. This is the usual vertical suc- 

 cession of the southern hemisphere 



But north of the line, when, for reasons which cannot 

 be discussed here, the south-east trade did not turn into 

 south-west, as might have been expected, the upper 

 currents continued to follow the succession of the 

 southern hemisphere, instead of conforming to the law of 

 the northern hemisphere. In the doldrums, which extended 

 from about 8^ to 13° N., the same rule obtained, and the 

 middle layer of cloud over some " cats' paws " of south- 

 east wind drove from the east. To show how difficult it 

 is to get cirrus observations in the tropics, I may mention 

 that this was the highest layer I was able to record during 

 this part of the voyage. 



In the north-east trade I only got one unsatisfactory 

 observation in 22° N., 19° W., which gave a middle layer 

 of north-north-east wind over an east-north-east surface 

 trade. This is the contrary of what might have been 

 expected. 



In the second section, between Teneriffe and Capetown, 

 the lower layers of cloud in the north-east trade — from 

 30° N. to the doldrums in 5° N. — invariably carne from 

 some point to the left of the surface wind, generally from 

 south or south-by-west. This is the usual vertical suc- 

 cession of the northern hemisphere. 



But as we entered the doldrums, in 5° N., a totally dif- 

 ferent wind-system became apparent. Over the oily calm 

 of that district I could just detect, through the universal 

 haze and gloom, a middle current from the east ; and when 

 a few hours later we picked up the south-west monsoon 

 of the Gulf of Guinea — here coming from south-by- 

 west — the low clouds came from south-east. This con- 



