FREED FROM THE ACTION OF GRAVITY. 21 
periments; but happily it may be obviated: this can evidently 
be effected by employing the two liquids only when they have 
already exerted upon one another all the action of which they are 
capable. The oil and the alcoholic mixture which I used are now 
inert with regard to one another, because having been employed 
a great number of times, they have had time to exercise the 
whole of their mutual action. Besides, it is easy, in a short 
time, to bring the two liquids to that state of relative neutrality, 
by agitating them together in order to divide the oil and thus 
to accelerate the action, then separating them by a suitable pro- 
cess. This operation requires some precautions, which we shall 
examine in § 24, in order not to interrupt the course of the me- 
moir by details which are not now indispensable. In all that 
follows we shall always suppose that two liquids thus prepared 
are employed. 
7. Another cause disturbs the equilibrium between the sphere 
of oil and the ambient liquid: this is the variations of tempera- 
ture, which alter the equality of the two densities; and the de- 
gree of sensibility of such a system in this respect would hardly 
be conceived. For example, when the vessel is carried into a 
room alittle warmer or colder than that in which it had been be- 
fore, the sphere soon falls in the first case and rises in the second. 
On the mere application of the hands to the outside of the ves- 
sel, it will be seen, after a few seconds, that the sphere begins to 
fall. 
We must be continually on our guard against these effects of 
temperature, otherwise they disturb the experiments. The fol- 
lowing is a recent instance which occurred to me. The oil and 
the alcoholic liquor were inclosed in different flasks, and the 
latter contained a very slight excess of alcohol. Having by 
chance carried these two flasks into a room warmer than that in 
which they had been, I first introduced into the mixture a cer- 
tain quantity of oil, which, by reason of the slight excess of 
alcohol, descended slowly to the bottom of the flask. A short 
time afterwards I poured in another quantity of oil, and I was 
surprised to see this on the contrary rise towards the upper part 
of the mixture. The reason of the singular difference was this: 
the alcoholic mixture inclosed in one of the flasks was very con- 
siderable in quantity relatively to the oil which the other con- 
tained. Now, at the first moment, the liquids, not having sensi- 
bly changed their temperature, maintained between them the 
