FREED FROM THE ACTION OF GRAVITY. 19 
oil slowly. Then if the alcoholic mixture is by chance ex- 
actly in the requisite proportions, the oil forms, at the extre- 
mity of the neck of the funnel, a sphere the volume of which 
increases gradually in proportion as we add this last liquid. 
When the sphere has attained the volume we desire, the neck 
of the funnel is withdrawn with caution; the sphere which 
adheres to it rises with it toward the surface of the liquor, 
and the oil which it still contains is added to the preceding. 
Lastly, when the sphere has nearly reached the surface of the 
alcoholic mixture, a little shake detaches it from the funnel. 
Ordinarily, however, the mixture has not so exactly the desired 
density. We then see, in general, several successive spheres of 
oil formed, which detaching themselves one after another from 
the mouth of the funnel, fall slowly to the bottom of the flask, 
or rise to the surface of the alcoholic liquor. In this case all 
these spheres should in the first place be united into one, which 
is easily done by the following means. We introduce into one 
of them the end of an iron wire: the adherence which the oil 
contracts with this metal then allows the sphere in question to 
be easily conducted in the ambient liquid, and to be led to join 
with a second sphere*; by continuing this treatment we soon 
succeed in uniting all. Then, according as the whole sphere 
shall remain at the bottom or on the surface of the liquor, add 
cautiously to the latter a certain quantity of water or of alcohol ; 
and, after having corked the flask, we next turn it several times 
slowly, and so as not to disunite the sphere of oil, until the mix- 
ture is well effected, which will take place when we no longer per- 
ceive any striz in the liquor on looking through it at a window. 
Lastly, the same operation is to be repeated until the sphere of 
oil is perfectly in equilibrium in the surrounding liquor. 
5. If the experiment has been made, as I have supposed, in a 
flask of the ordinary form, that is to say cylindrical, the mass of 
oil does not however appear exactly spherical; it is widened in 
the horizontal direction; but this is only an optical illusion, 
attributable to the form of the flask: the latter, with the liquor 
which it contains, acts in the manner of a cylindrical lens whose 
* In order thus to compel two spheres to unite, it does not suffice to put 
them in contact with one another: they might touch for a long time without 
mingling into one; one would say that they are enveloped in a resisting pelli- 
cule which opposes their union. It is also necessary, therefore, to introduce 
the extremity of the metallic wire into the second sphere, as if we wished to 
break the partition which separates the two masses: the union is then effected 
immediately. I shall revert to these phenomena hereafter. 
Gz? 
