atmospheric mass. Now at such an akitude the pressure is, naturally, 

 no more than one-tenth of an atmosphere. In other words, at this 

 height the pressure of the air is no more than a tenth of that we ex- 

 perience at sea-level. As the lifting force of a balloon is proportional 

 to the density of the air displaced, as Archimedes would have already 

 told us, I had thus to construct a particularly large and light balloon, 

 so that it could carry observers, instruments and the airtight cabin. 



I spare the reader the calculations I made : I had to have an envelope 

 of 223,560 cu. ft., of 1 14 ft. in diameter, made of a material of the least 

 possible weight. Here arises the principal difficulty in the construction 

 of stratospheric balloons: a balloon of this volume, completely 

 inflated with hydrogen, would have, at its take-off, a static lift of nearly 

 16 tons. To resist this force, material and net would have to be extra- 

 ordinarily strong, and thus heavy — so heavy that the balloon would 

 never reach 10 miles, where a cubic yard of hydrogen supports only 

 one-tenth as much as it does down here. To permit the use of a light 

 envelope, then, it was necessary to introduce into our balloon, at the 

 moment of take-off, only a small part of the gas that it could contain, 

 one-fifth of its maximum volume. During the ascent this gas would 

 expand under the effect of the decrease of atmospheric pressure and 

 only in the stratosphere would the envelope take its spherical form. 



Which of my readers has been present at the rigging of a spherical 

 balloon ? The envelope is spread out on the ground, like a cast-net. 

 Upon it the net is disposed. The gas is introduced. The envelope 

 dilates and lifts up the net, which is held (and stretched) by bags of 

 ballast. As the volume of the envelope increases, the bags are taken 

 from mesh to mesh to be hooked on lower down. During this whole 

 operation care must be taken that the folds in the expanding envelope 

 open out completely, without being caught in the folds of the net. 

 When the envelope has become spherical and has attained the desired 

 height, the ropes attached to the net are affixed to the hoop and the 

 balloon is prepared for the ascent. All this is accepted practice. 



But our FNRS was to receive, at the beginning, only a small part 

 of the gas which would later inflate it entirely. It was thus only the 

 upper portion which would contain gas, the rest of the envelope 

 remaining empty and hanging in great loose folds which would be 

 progressively filled during the ascent. In these circumstances what was 

 to be done to avoid the accidents arising from folds partially retained 

 in the net ? We could not count upon a procession of guardian angels 



[4] 



