22 PLATEAU ON THE FIGURE OF A LIQUID MASS 
same relation of density ; but after a short time the oil, by rea- 
son of its small volume, having become warmer than the alco- 
holic mixture, had thus become relatively lighter. The warmth 
of the hand which held the flask in pouring out the oil must 
have also contributed to the effect in question. 
8. Now let us suppose a fine sphere of oil in permanent equi- 
librium in the surrounding liquid, and let us endeayour to sub- 
mit it to other forces than its own attractions. 
The first idea which presents itself is to try the action of cen- 
trifugal force. For this purpose it is necessary to impress on 
the sphere of oil a movement of rotation around one of its 
diameters, and which is effected by introducing into this sphere 
a small metallic disc, which is made to turn upon itself by means 
of an axis which traverses it perpendicularly. This disc carries 
the oil with it by its adherence, and the whole mass of this liquid 
takes a movement of rotation. 
Before explaining the effects which result from this movement, 
I shall describe in detail the apparatus I have employed,—an 
apparatus by the aid of which all the experiments succeeded 
perfectly and with the greatest facility: it is represented in fig. 1. 
Fig. 1. 
The vessel is with plane sides, formed of rectangular plates of 
glass set in an iron frame; the sides are each 25 centimetres 
broad and 20 high. The small disc and its axis are also of 
