410 LETTERS FROM THE ANDREE PARTY, 



and I made a ruiuute calculation of how long the gas would remain in 

 the bag, taking into consideration its slow escape through the minute 

 interstices which we found it impossible to close up. Our calculation 

 was that the balloon would remain afloat, barring accidents, not longer 

 than fifteen days. The "Ornen" probably was the tightest balloon 

 bag ever made, but we could not close up some of the holes. I invented 

 a process for detecting the escape of the gas. After the balloon was 

 inflated we spread long strips of sheeting saturated with acetate of 

 lead over the top. The confined hydrogen sulphide as it escaped 

 would, on coming iu contact with the sheeting, cause the latter to 

 become discolored. Thus the exact location of every hole could be 

 ascertained. 



Even so, although we varnished and revaruished the silk inside and 

 out, we could not prevent the almost imperceptible holes. The great- 

 est trouDle was in the seams where the sections of the great bag were 

 joined. The finest needle hole was sure to show a leak, even after the 

 stitching had been done as neatly as j)ossible. It may not be known 

 that the successive varnishings could be done satisfactorily only on the 

 upper part of the bag, against which the greatest outward pressure of 

 the confined gas was exerted. Of course we would have done the 

 lower part of the balloon more thoroughly, but we had not time. The 

 aeronauts were eager to sail with the first good south wind, and we 

 had to let them go. Strindberg had made some exjDeriments to reduce 

 the outflow of the gas, but they were without success. The constant 

 smearing on of the gutta-percha which we used was really the best we 

 could do. Our estimate of fifteen days' duration for the gas, small as 

 it was, would provide for its retention twice the length of time that 

 gas has ever before been confined in a balloon. I believe that no bal- 

 loon heretofore has floated longer than a week. 



Andree was handicapped at the start by the loss of two-thirds of 

 his drag ropes, upon which he depended to steer his balloon. The acci- 

 dent was a curious one, and while it could not have been foreseen, yet 

 the conditions under which it happened might have been avoided if 

 diflerent arrangements had been made. The drag ropes of the balloon, 

 which were about 1,000 feet long, were in three sections, and were 

 joined together by metal screw couplings. The couplings could be 

 screwed apart, the inference being, I suppose, that if Andree wanted 

 to do so he could unscrew and cast off any part of the drag rope. Why 

 he could not just as easily have cut them apart 1 do not know. Now, 

 when the balloon was ready to start, the drag ropes attached to the 

 lower side of the basket were allowed to trail up over the top edge of 

 the balloon house and down again to the beach along which they were 

 trailed, so as to be clear of all obstructions and ready to follow the 

 balloon out to sea when it rose out of its nest. 



But the heavy weight of the ropes defeated this purijose. The part 

 of the ropes which lay outside on the beach offered an immense friction, 



