during the sunset descent (caused when the superheated helium lost the sun's heating effect). A further descent of 5 mb (500 feet at this level) was required to start the flow of ballast. By this time, the balloon had lost considerable lift and in exchange had acquired a downward velocity of about 120 feet per minute. To check this descent a ballast flow was required for about 40 minutes. During the next hour the balloon was buoyant and climbing back to the seal-off pressure of the automatic ballast valve. The inefficiency of this valve system is demonstrated by the ballast which was lost after the balloon had regained its buoyancy and had begun to rise. More ballast was wasted than was required to check the descent. Indeed, the entire 3000 grams available was expended at this time, according to the evidence of the ballast meter.
On this flight there was no apparent change in the activation pressure of the automatic ballast aneroid between the times when ballast flow began and ended. This indicates that the entrapped air had not experienced any significant temperature change during the two hours of ballast operation.