348 PROCEEDINGS OF THE AMERICAN ACADEMY. 



upon their initial distension. A parachute, covering the top of the 

 balloon, moderates the fall and it reaches the ground comparatively near 

 the point of departure. The times at which the balloon left the ground 

 and returned to it being automatically recorded and these places being 

 known, the average direction and velocity of the balloon can be calcu- 

 lated, although its velocity is usually greater over the upper portion of its 

 course. In some cases two balloons, unequally filled and coupled 

 tandem, are employed, and, as only one balloon bursts, the other is 

 borne slowly to the ground, where it continues to float and serves to at- 

 tract persons to the spot. If the balloons are not badly torn they 

 may be mended and used again, but, since the rubber has been perma- 

 nently stretched, its elasticity is impaired and consequently the balloons 

 will not rise as high as before. The quality of the rubber also deterio- 

 rates rapidly, especially when exposed to light and heat. The recording 

 instruments, which were furnished by M. Teisserenc de Bort of Paris, 

 consist of a metallic thermometer and a Bourdon barometric tube which 

 record upon a smoked cylinder turned by clockwork. The smoked 

 sheet of aluminum is afterwards removed from the cylinder and the rec- 

 ord made permanent by flowing shellac over it. The temperature and 

 barometric pressure are recorded simultaneously, and from them the cor- 

 responding curve of height may be computed with tolerable accuracy by 

 Laplace's formula. Although the instruments had been calibrated in 

 Paris, they were again tested before each ascension for barometric pres- 

 sure under the receiver of an air-pump, in which the natural rate of 

 change of pressure was imitated as nearly as possible, and for temper- 

 ature in a bath of alcohol cooled with solidified carbon dioxide. Each 

 instrument is contained in a mica case and further protected against 

 rough usage by a wicker basket suitably screened against the sun's 

 rays, the whole weighing about a pound and a half. It is suspended 

 from the parachute at a considerable distance below the balloon and 

 directions printed on a water-proof envelope request the finder to pack 

 the instrument carefully in a box and return it either to St. Louis or to 

 Blue Hill, a reward being promised for the service. A card contained 

 in the envelope, after being filled out with the finder's name and address 

 and the time of finding, is to be mailed immediately to the sender. The 

 fact that all but one of 23 balloons and instruments despatched were 

 found and returned shows the efficiency of these arrangements as well 

 as the excellent topographical situation of St. Louis for this work. 



Owing to difficulties in obtaining hydrogen gas on the Exposition 

 Grounds, the experiments were not begun until the middle of Septem- 



