108 Prof. Wartmann on the Polarization of Atmospheric Heat. 



vacuum obtained by tbis process, arises from tbe difficulty of pre- 

 paring carbonic acid gas perfectly free from air. Tbis may be 

 very nearly overcome by adopting precautions wbich are well 

 known to practical cbemists. When an extreme exhaustion is 

 required, the gas-holder shoidd be filled with recently boiled 

 water, and the first portions of carbonic acid that are collected 

 in it should be allowed to escape. 



The substitution of phosphoric for sulphuric acid would re- 

 move the possibility of either aqueous or acid vapours being 

 present even in the smallest amount, but such a refinement will 

 rarely be found necessary. 



In the experiment just described, the theoretical residue of 

 air would be — ^dth part of the entire quantity in the receiver, 

 which would cause a depression of ^dth of an inch. This result 

 must have been nearly realized. If the exhaustion had been 

 carried at each time to 0*2 inch, the residue by theory would 

 have been only • * 'th part. But the experimental results will 

 not continue to keep pace with such small magnitudes. 



Queen's College, Belfast, 

 January 7, 1851. 



XVI. On the Polarization of Atmospheric Heat. By Elie Wart- 

 mann, Professor of Natural Philosophy in the Academy of 

 Geneva*. 



THE observations of M. Arago and Sir David Brewster have 

 long since established, that the light by which our atmo- 

 sphere is illuminated is polarized in certain directions. It might 

 be supposed from analogy, that the heat proceeding from the 

 same source is endowed with similar properties ; the following 

 experiments place this supposition beyond the pale of doubt. 



The means of polarizing a ray of heat, without greatly dimi- 

 nishing its intensity, are less perfectly known than those of po- 

 larizing a ray of light, and the result is a corresponding infe- 

 riority in the exactitude with which the calorific ray can be ana- 

 lysed. In the use of the thermo-electric pile, the experimenter 

 must be on his guard against numerous sources of error. The 

 blackened face of the instrument radiates into space, and is 

 cooled to a degree which depends partly upon the transparency 

 of the air, partly upon its temperature. The other face, 

 although protected by a closed tube, is not entirely free from the 

 influence of conduction in prolonged experiments. The thermo- 

 metric state of the atmosphere changes capriciously every moment, 

 owing to the unequal mixture of the ascending and descending 



* From the Bibliotheque Universelle, October 1851. 



