UPPER ATMOSPHERE 21 



meteorological kites, with the necessary wire and winch 

 and special meteorographs ; for ascents can easily be 

 made from a ship, which can be manoeuvred to help 

 the wind in raising the kite ; but no direct investigation 

 of the upper air is practicable on a vessel which cannot 

 deviate from her course. A good deal can be done, 

 however, to determine the direction of the upper 

 air currents by letting off small balloons (pilot balloons), 

 and measuring the angle of altitude of the liberated 

 balloon by means of a sextant at uniform intervals of 

 time, and the azimuth at the same intervals by means 

 of a compass provided with suitable sights. The 

 altitude angles give the distance of the balloon from 

 the ship on the assumption that the rate of ascent is 

 uniform, the time intervals thus giving the height 

 above the sea at each observation. The rate of 

 ascent for balloons of a given size, material, and 

 amount of hydrogen, must be determined by measure- 

 ments on land, by two theodolites at the ends of a 

 measured base. The azimuths, taken in conjunction 

 with the ship's course and speed, show the direction 

 of drift of the balloon as it rises — that is, the movement 

 of the wind at the given heights. 



Without the use of any instruments the character 

 and movements of the upper clouds are an indication 

 of the winds in the higher regions of the atmosphere ; 

 and, when travelling in places where volcanic eruptions 

 may occur, the observer ought to take particular note 

 of the form, angular height, and direction of drift of 

 any column of steam or smoke which may be seen to be 

 thrown up by a volcano. It may be recalled that the 

 dust of the eruption of Krakatoa, in 1883, which was 

 projected to an observed height of twenty miles, 

 was carried by the upper currents of air, not only all 

 round, but all over the world. In all observations 



