679 



from 16 h to 21**, are greater than the real 

 variations, because at those epoch as the ba- 

 rometer and thermometer rise and sink to- 

 gether. 



The same thing has happened in respect to the 

 horary variations of the barometer, as takes place 

 with respect to a great number of important phe- 

 nomena, which the history of physical discoveries 

 displays in the first instance, that are either 

 vaguely perceived, or carefully examined, but 

 published by insulated observers, who enjoy little 

 celebrity. These phenomena remain forgotten if 

 the learned, or the academies, which in every 

 age exert a great influence on the progress of the 

 sciences, have not made them an object of their 

 researches. When, afterwards, by the union of 

 several observers known by other labours, or by 

 a more complete discussion of the phenomena, 

 doubts are dissipated, things are then eagerly 

 recognized as anciently known, which it is no 

 longer permitted to neglect as ill-observed. A 

 learned man, father Cotte, who has rendered 

 essential services to meteorology, attributed, in 

 1774, notwithstanding the uniform testimony 

 of so many travellers who had visited the tro- 

 pics, the regularity of the horary variations to 

 the imperfection of the barometers, that is, 

 to a small quantity of air contained in the 

 void of Torricelli, and susceptible of being di- 

 lated and condensed by the increasing and de- 



