4 Transactions of the Royal Microscopical Society. 
sions of the vase and the mean length of path of the molecules, 
descends below and the maximum effect of the pressure of the 
gas upon the disks takes place when this ratio becomes equal to 
twenty. Admitting the preceding data as established, which 
accord very well with results obtained by Sir. Crookes, we are 
therefore justified in attributing the oscillatory motion of little gas- 
bubbles imprisoned in a liquid, to the velocity of translation of the 
molecules, varying in intensity and direction from one to another, 
and giving rise at the same instant, by default of the total commu- 
nication of pressure, to unequal propulsions upon the different points 
of the envelope. The tension of the gas, the mass of the molecules, 
and the coefficient of resistance of the surrounding liquid, will 
determine, in each particular- case, the conditions of volume for 
which the Brownian motion ceases or is produced. 
The bubbles of the liquid cavities in quartz are very probably 
bubbles of vapour, produced when the strata were formed by eva- 
poration, on the cooling and the retreating of the liquid. The 
Brownian motion of these vapour bubbles, and the molecular motions 
of the little gas-bubbles, are due to the same cause. It is true, how- 
ever, that, according to the mechanical theory of heat, the state of 
equilibrium of a mass of vapour “ is not a state of repose, in which 
the vaporation has ceased ; but a state in which there is a continual 
evaporation, and a condensation equally strong, which counteract 
one another.” * When a molecule of vapour strikes the liquid, the 
latter “ does not generally repel it, but retains it, and assimilates it 
by the attraction which the other molecules immediately exert upon 
that which approaches.” f This phenomenon does not take place 
in gases. But at the same time as the condensation of which we 
have just spoken is produced, an equivalent evaporation is deve- 
loped at the same place. This simultaneous evaporation makes 
up for the losses which the liquefaction causes in the vapour, and 
thus maintains the tension and density of this latter in an invariable 
state. 
Moreover, the pressures determined by the collision of the mole- 
cules of vapour, and the evaporation accompanying it on the dif- 
ferent points of the surrounding liquid, are evidently equal to those 
that an elastic gas would produce in the same conditions of volume, 
density, tension, and temperature. From this I conclude, that the 
cause producing the Brownian motions in small gaseous bubbles, is 
also that which produces these motions in the vapour bubbles of 
liquid cavities. This conclusion is by no means weakened, in the 
case of bubbles in quartz, from the consideration of the ratio between 
the dimensions of the surrounding liquid, and the mean path of the 
molecules. For, in the assimilation of steam with the atmospheric 
* 4 The'orie Mecanique de la Chaleur,’ 2* partie, p. 195. 
t Ibid. p. 194. 
