Temperature and Pressure in Balloon Ascents. 215 



without the water in the water-case sinking below 0°. In any 

 case, for greater safety, a bottle full of quicklime, or of caustic 

 potash, may be taken; or indeed warm water, which can be 

 used as ballast; and the temperature of the water in the case 

 may be raised at will by introducing the substances by the tubu- 

 lure 0. 



This ice-apparatus will be placed inside the boat ; it occupies 

 no great space, and its weight will be no hindrance for a balloon 

 ascent ; for it constitutes part of the ballast, which may be ad- 

 vantageously used to moderate the descent, by allowing the water 

 to flow out of the water-case by the stopcock R. 



Suppose the two systems conveniently arranged, all the stopcocks 

 being open. The balloon is ascending; to determine at a given 

 moment the temperature and pressure of any atmospheric layer, 

 the observer closes simultaneously the stopcocks of the tubes 

 No. 1 of the two svstems. He will do the same at all the sta- 

 tions at which he desires to make an observation. To this sim- 

 ple operation all his work is limited. 



The moment of the determination is not immaterial; the 

 balloon ought not to be ascending or descending too rapidly; 

 this is easily recognized by the motion of a sheet of paper 

 attached by a thread to one of the cords of the boat. When the 

 balloon ascends, the sheet sinks and sticks to the cordage ; when 

 it descends, on the contrary, the sheet rises, and the velocity of 

 the descent is estimated by the amount of the divergence. The 

 most favourable times for observing are those when the balloon 

 is almost in equilibrium in the atmospheric layer, and makes 

 oscillations, the existence of which is seen by the motion of the 

 sheet of paper. 



After the descent, the apparatus are brought to the laboratory 

 of the physicist ; and in a very short time he can accurately de- 

 termine both the temperatures and the atmospheric pressures. 

 For this purpose he uses a mercury manometer (fig. 9) like those 

 which I frequently use for air-thermometers. The capillary tube 

 bed, in which the tube a b terminates, is provided with a steel 

 socket d, which exactly fits by means of a screw-press (fig. 5) on 

 each of the stopcocks r of the tubes of the barometric system, and 

 of the thermometric system, used in the balloon ascent. On 

 the tube ab are two marks, one at a, the other at /3; and by 

 weighing the mercury which fills them, the capacities v and v' of 

 the tube a b comprised between the levels «, /3, and the end of 

 the capillary tube bed are exactly determined. 



Let 



x be the unknown temperature of the layer of air in which 

 were the tubes No. 1 of the two systems at the time of 

 closing. 



