48 M. Savart on the influence of different Media, Sfc. 



not be increased, but diminished, as the approach to a vacuum 

 in the space above the mercury would tend not to expand, 

 but to contract the bulb. It is generally admitted, that no 

 perfect vacuum has ever been produced. However this may 

 be, we are certain that, in the most carefully constructed ther- 

 mometers, some air is left in the interior of the instrument ; for 

 the means employed for introducing the mercury, and expel- 

 ling the air, are not sufficient for wholly abstracting the lat- 

 ter from the inner surface of the glass. Let it then be admit- 

 ted, that some air, more or less, has been left in the bulb and 

 stem of the instrument ; in the course of years, this air will be 

 decomposed by the mercury, the oxygen, at least, will be ab- 

 sorbed, and, in becoming solid, will have its bulk greatly di- 

 minished. In this way, the lowering of the mercury from its 

 original height may be accounted for ; — but even before the in- 

 cluded air has been decomposed by the mercury, some of it 

 that has adhered to the inner surface of the bulb, and that 

 part of the stem that was not left empty, will escape to the 

 upper part of the tube during the frequent expansions and 

 contractions of the mercury, and thus occasion a slight differ- 

 ence in the height of the fluid, from what it was when the in- 

 strument was constructed. In every instance the observed 

 change has been small; and if the above explanation be the 

 true one, the diminished height will, in each instance, be in 

 exact proportion to the quantity of air that has been left be- 

 tween the surfaces of the glass and the mercury. Hence the 

 most obvious and certain preventative would be, to allow a 

 considerable space of time to elapse between the construction 

 of the glass part of the instrument, and the adaptation of its 

 scale. 



Akt. IX. — On the Influence exerted by different Media on 

 the number of Vibrations of Solid Bodies. By M. Felix 

 Savart.* 



Attempts have frequently been made to determine the num- 

 ber of vibrations of solid bodies, when made to sound suoces- 



* Translated from the Ann. de Chimie, <$-c. Nov. 1825, p. 264-269. 



