AEROSTATIC VOYAGE. 233 



--The courfe of the hygrometer was very lingular. At the Unfteadycourfe 

 furf'ace of the earth it was only at 57.5°, while at the height JJ^ h ^ VQ ' 

 of 3032 metres it marked 62° : from this point it was conti- 

 nually falling to the height of 5267 metres, where it indicated 

 no more than 27.5 <? ; and from thence to the height of 6S84 

 metres, it rofe gradually to 34.5°. U, by thefe refults, it is 

 ivifhed to determine the law of the quantity of water diilblved 

 in the air at different elevations, it is clear that attention 

 muff, be paid to the temperature ; and on joining this confider- 

 ation with it, it will be feen that it follows an extremely de- 

 creafing_ progreliion. 



If now the magnetic ofcillations are confidered, it will be Magnetic ofcll- 



oblerved that the time for ten ofcillations, made at different latio " s a ' difter - 

 i • i • r ent heights, 



heignts, is fometimes above and fometimes below 42. 16 , 



which they require upon the earth. By taking the mean of 



all the ofcillations made in the atmofphere, ten ofcillations 



would require 42. 20", a quantity which differs very little 



from the preceding; but on coniidering only the laft obfer- 



vations which were made at the greateft heights, the time 



for ten ofcillations would be a little lefs than 42.16°, which 



would indicate, on the contrary, that the magnetic power 



had been a little augmented. Without meaning to draw any 



coniequence from this apparent flight increafe, which may 



very probably arife from the errors to which experiments of 



this description are liable, I conclude that the total of the The magnetic 



refults I now offer, confirms and extends the fact which M. Peer's not fen- 



Biot and myfelf had obferved, and which, as well as the rne greateftele- 



univerfal gravitation, proves that the magnetic power does vations » 



not experience fenfible variations at the greatefl heights to 



which we can afcend. 



The confequence we have drawn from our experiments, Probable objee- 



will appear a little too precipitate to thofe who recoil*!® that tions , t& lh '' 

 , . , . ..... .conclufion re- 



we were unable to make experiments on the inclination off ute d € 



the touched needle. But if it be obferved that the power 



which caufes an horizontal needle to ofcillale, is neceflarily 



dependent upon the intenfity and direction of the magnetic 



power itfelf, and that it is reprefented by the co-fine of the 



angle of inclination of this latter power, they cannot avoid 



concluding with us, that, fince the horizontal power has not 



varied, neither can the magnetic power have varied, unlefs - 



bhey fuppofe that the magnetic power can have varied pre- 



( ci fely 



