and on Astronomical Refractions, 



219 



Temperature cent. 

 Temperature observed by Gay Lussac — -•— — ••*■— 



Temperature calculated . 



The abscissa represents the temperature in degrees of the centigrade thermometer, and 

 the ordinate the height in miles. 



At first the decrements of temperature are nearly equal for 

 equal increments of altitude. These observations by no means 

 furnish so good a criterion of the accuracy of my formula as the 

 observations which have been made of the temperature of steam 

 and other vapours. The determination of the constants y and E 

 for the atmosphere must be repeated at some future time ; for it is 

 obvious ^that no great reliance can be placed upon the extreme 

 precision of the values now obtained * until other ascents have 

 been made, and many similar observations have been compared to- 

 gether. We may then hope to obtain constants accurately apper- 

 taining to a mean state of the atmosphere, and the variations which 

 take place in their values corresponding to fluctuations of the tem- 

 perature ; the pressure and the humidity of the atmosphere at the 

 earth's surface may then be investigated. M. Biot has suggested 

 that balloons furnished with self-registering instruments should 

 be moored over each of the principal observatories of Europe. This 

 plan appears to me subject to great difficulties. The weight of the 

 line attached will diminish the buoyancy, so that I apprehend it 

 will be found impossible to send up a balloon so fastened to any 

 considerable altitude. The escape of gas will, I imagine, render 

 it very difficult to maintain the balloon at the same height for any 

 length of time. The height of the balloon will also be subject to 

 great variations from the tension of the line changing with the force 

 and direction of the wind. I am disposed to attach much greater 



* Dulong found, for atmospheric air, perfectly dry, y 

 torn, ii. p. 646. 



1*421. See Poisson, Mdc,, 



