18 SCIENCE PROGRESS 



absence of strong winds or cloud. Moreover single theodolites 

 were used, and in calculating the height of the balloon it was 

 assumed that the rate of ascent was the same at great and at 

 moderate altitudes, which has not yet been certainly proved true. 

 Rate of Ascent of Pilot Balloons (C. J. P. Cave and J. S. 

 Dines, Q. J. Roy. Met. Soc, Oct. 191 9). In a previous paper 

 {Q. J. R. Met. Soc, vol. 39, 1913) Dines found, as a result of 

 experiments in a balloon shed at Farnborough, the following 

 empirical formula connecting the rising velocity in still air 

 with the dead weight and free lift of a balloon : 



where V = rising velocity in metres per min. 

 L = free lift in grammes. 

 W = weight of balloon in grammes, 

 and ^ is a constant for a particular balloon. 



It should be noticed that V is directly proportional to this 

 quantity q. This formula is now in general use in this country. 

 The question of the rate of ascent of a balloon is, however, 

 not quite so simple as the above formula implies, and in any 

 case the introduction of small lanterns suspended beneath the 

 balloon for use in locating its position at night made further 

 experiments desirable. These were carried out in the Royal 

 Albert Hall. 



A difficulty encountered throughout the work was the 

 fluctuation in the speed of the same or similar balloons from 

 time to time. A series of observations would yield values of 

 q which varied but slightly among themselves, yet which 

 differed b}'- as much as 5 per cent, from the mean of a longer 

 series. Since no explanation could be found to account for 

 this, a large number of observations had to be made when 

 investigating any particular point. The first experiments 

 were on the effect of loading. By hanging a small weight 

 to the neck of a balloon, the rate of ascent was in certain cases 

 increased, e.g. with a 30-grm. balloon having 120-grms. lift, 

 the variations in q for different loads were as follows : 



Load o 5 10 15 20 30 40 grms. 



q 83-6 87-0 87-4 87-0 84-9 84-7 84-2 



When loaded, the balloon was observed to ascend in a much 

 less zigzag course than when in the unloaded state, and it 

 was doubtless for this reason that the ascensional velocity 

 was greater up to a certain point. 



The next experiments were to determine whether the cotton 

 which was normally attached to the balloon to assist in measur- 



