reveal themselves bit by bit: a summit, then another: at last, all the 

 snowy chains of the Bavarian Alps and the Tyrol, which we are 

 approaching gradually. 



In spite of the splendour of the spectacle, we took precautions. We 

 threw out over a hundred pounds of ballast, which caused us to rise 

 some hundreds of yards. 



We soon made a very unpleasant discovery: the rope which 

 controlled the valve was not working. It was tangled with the supple- 

 mentary rope which was affixed at the moment of departure. Now, if 

 we could not open the valve, we could not let the gas escape, to begin 

 the descent. Instead of obeying us, the balloon would go down only 

 when external conditions permitted it, that is to say, when it grew 

 colder at sunset. Where should we be then ^ Over the land ? Or above 

 the Adriatic ? 



As it descended, the balloon would grow longer : the rope operating 

 the stopped-up valve would therefore be stretched out, and would 

 open the valve, accelerating our descent more than we wished. 



However, to carry out our programme and reach that altitude where 

 the pressure is only one-tenth of an atmosphere, we threw out more 

 ballast and soon we saw a difference on our barometer between the 

 two meniscuses of 2-992 in. exactly. Being used to seeing in the 

 laboratory, on our barometers, columns of mercury of 29-92 in. we 

 had a curious sensation when we read a barometric height reduced to 

 one-tenth of what we call its normal value. 



We should have been perfectly happy if it had not been for this 

 incident of the valve. The future was uncertain. What were we to 

 do.'^ We decided not to throw out any more ballast, partly to 

 shorten our trip, and partly, also, to be able to dispose of what re- 

 mained at the moment of landing. Then we decided to pack up the 

 instruments. If the balloon, as it drew out in length, itself pulled open 

 the valve and thus occasioned too sudden a landing, we had to take 

 precautions against being injured by loose objects. 



We tried once more to open the valve by turning the windlass 

 winch around which the cable was wound, by means of a crank placed 

 inside the cabin. But the cable broke clean off, which definitely put 

 at an end any hope of controlling the balloon. 



There we were, prisoners of the stratosphere. Fortunately we had 

 at our disposal a good reserve of oxygen, and of alkali, which is used 

 to absorb the carbon dioxide produced by our breathing. Although 



