Chemical Analysis of the Atmosphere. 283 



(by pressing it first tbrough a tube containing moist cotton,) without 

 finding the second tube altered in weight. 



The increase of weight of the hygrometric tube, in causing the 

 discharge of 12972.5 grammes of water, is found, under ordinary 

 circumstances, included within the limits of sixty to one hundred 

 milligrammes, an amount apparently considerable enough to be but 

 Htt^e influenced by errors of observation. On the 26th of Dec. 

 1830, during the finest weather, at a temperature of 16° cent. (=€0° 

 Fah.) I obtained thirteen milligrammes from the same volume of air. 



Although this method of estimating the moisture of the atmosphere 

 always requires an experiment, and for which common hygrometers 

 cannot be substituted for the instrument now described, it appears to 

 me that this is the only one which can give a direct result, and that 

 it must on that account be useful in regulating others. I think never- 

 theless, it may be used successfully in travelling. A tin vessel, of 

 convenient dimensions, (from eight to ten thousand cubic centime- 

 ters) would be easily transported, as well as tubes containing amian- 

 thus moistened with sulphuric acid, and weighed beforehand, the ba- 

 rometer and thermometer always making a part of the equipment of 

 a scientific traveller. 



To those who could not conveniently obtain a vessel expressly 

 adapted to this purpose, I would propose the use of a common bot- 

 tle, thus — fill the bottle A, (Fig. 2.) with water, which may be drawn 

 off by the syphon a6, and measured by means of a vessel of known 

 capacity. To tlie tube c adapt the hygrometric tube in the common 

 manner. With this apparatus, which may be constructed wherever 

 bottles can be found, and a small portable measuring glass, with the 

 weighed tubes, it is easy to make experiments of sufficient exactness. 

 Care must of course be taken to guard against the influence of hu- 

 midity arising from the flow of the water, the decanting from the 

 measuring glass into the bottle, Sic. which will never be difficult. 

 With a fixed apparatus, it would be advisable to perform the experi- 

 ment near a window, through an opening of whicli the hygrometric 

 tube should pass into the open air. 



II. Determination of Carbonic Acid. 



The valuation of the carbonic acid in the atmosphere has been 

 less studied than that of the othev elements, either because it pre- 

 sents greater obstacles or is judged to be of less interest. All that 

 is positively known, we owe to Theod. de Saussure, whose memoir^ 



