42 PLATEAU ON THE FIGURE OF A LIQUID MASS 
there is too little density, we correct it with alcohol at 16 de- 
grees. In this latter case we must not use pure water, because 
this, when it mixes with the prepared alcoholic liquor, produces 
in it a cloudiness more or less decided. 
The various trials I have made relatively to the above process, 
have led me to ascertain that the two liquids, when they have 
not been submitted to this preparation, are both modified by 
their mutual contact. The alcoholic liquid dissolves some 
oil, and this in its turn probably dissolves some alcohol. It 
is especially from the modification which the oil undergoes that 
its great diminution of relative density results ($6). Now, when 
the oil thus modified remains exposed to the air, it passes again 
gradually to the state of fresh oil, and resumes its former den- 
sity. It is partly to avoid this that I have recommended the 
funnels which enclose the filters to be kept constantly covered, 
and the oil to be kept in a flask perfectly corked. As for the 
alcoholic mixture, it is evident that this last precaution is equally 
necessary. 
25. Before I conclude, I must forewarn those persons who 
may wish to repeat my experiments, of two effects which some- 
times occur, and which cause disturbance in the operations if 
the experimenter does not know the means of preventing or de- 
stroying them. 
When some oil is introduced into a mixture containing an 
excess of alcohol, it happens sometimes that the mass which 
has sunk to the bottom of the vessel contracts adherence with 
this bottom and spreads itself out more or less on its surface. 
There is then no means of removing it entire ; but the spreading 
of the adhesion may be prevented by contriving that the bottom 
of the vessel should be occupied by a layer of a mixture more 
dense than the oil (§ 9). 
The second effect to which I allude is presented in the in- 
verse case, that is to say, when the sphere of oil, instead of 
reaching the bottom of the vessel, rises, on the contrary, to the 
surface of the alcoholic liquor, either because this liquor contains 
too little alcohol, or on account of a lowering of temperature, or 
because we have not been able to use prepared oil. When this 
happens, the mass flattens at first more or less at the surface of 
the mixture, as if this last opposed a resistance to it ; then, after 
some time, it makes its way through, and then presents a por- 
tion of plane surface, more or less extended, on the level with 
