through the jftmosphere. ' 323 



^Dl be deemed very accurate, on account of the uncertainty of 

 the present theories of resistance, but will serve to give us a 

 general idea of the effect likely to be produced. 



The first experiment was performed on April IS, 1814. The 

 diameter of the balloon was 4 "5 feet, and the plane was con- 

 structed in the form of an ellipse of which the diameters were 

 4*5 and ^y2>'c>. The plane being elliptical instead of square 

 «ould not possibly make anv perceptible difterence in the effect, 

 which whoilv depends on the surface. 



The balloon was inflated in a building of considerable height, 

 and there suffered to ascend. The experiment was attended 

 with the utmost success : for the balloon, instead of ascending 

 perpendicularly, actually moved at an angle o/ 45 ' till it reached 

 the ceiling. The power of the balloon being soon exiiausted, it 

 descended along the very track it had described whilst ascending, 

 and returned to the s])ot whence it had set out. Here the re- 

 sult j>reseHts a striking coincidence with the preceding calcula- 

 tion. 



This experiment was afterwards frequently repeated with a 

 square plane instead of an elliptical one, as I found the former 

 ■shape better adapted for furling- There was also added the 

 triangular sail £ (fig. \) before mentioned, which, l>y increasing 

 the resistance on the side to which it may be moved, is calculated 

 to answer every purpose of a rudder. The experiment was now 

 equally successful, the balloon turning to one side or the -other, 

 according to the position of the small sail. 



The same experiment was jierformed in the open air for the 

 first time, Octol^er 12, 1814. The balloon's diameter was 3*3 

 feet, and the length of the side of the square plane 4*25 feet. 

 The triangiriar sail was fixed to one side, for the following rea- 

 son : The balloon rising in the open air is carried by the wind 

 at the same time that it is acted upon by the attac'ied plane, 

 and it would, therefore, be difficult for an observer to determine 

 the effect produced by the apparatus. The triangular sail serves 

 as a test for this })urpose ; since, if the great plane did not act, 

 the former would feel no resistance, and no particular effect be 

 produced on the balloon. But if the great plane did act, the 

 small bail would constantly experience resistance, and have the 

 effect of making the balloon ascend in a spiral direction. The 

 balloon with its appendages was launched, and after a few se- 

 conds began to describe the expected spirr.i, which it continued 

 to do until lost in the clouds. 



On October 2, ISI 5, another experiment of the same nature 



was made with a balloon six feet diameter, having a square plane 



whose side was 7*5 feet, and a triangular rudder in proportion. 



The effect produced during the ascent was perfectly similar to 



X 2 that 



