, WITHDRAWN FROM THE ACTION OF GRAVITY. 215 



gradually of itself assumes such a position that the axis of the disc traverses it 

 diametrically. This effect is evidently owing to the attractive action of this 

 axis, or rather of the coating of oil with which it has been moistened — an 

 action which tends to operate in a symmetrical manner all around it, and thus 

 brings the entii'e sphere of oil into a position symmetrical with respect to this 

 same axis. Now it will be seen that the centre of the sphere tending, on the 

 one hand, to remain at the height of that of the disc, on account of the super- 

 position of the alcoholic layers of unequal density, and, on the other hand, to 

 place itself in the axis of the disc, on account of the symmetry of the 

 attractive actions exerted by the latter upon the oil, the centre of the sphere 

 and that of the disc will coincide, and will tbus remain in a fixed position. 

 Only the sphere will then be slightly elongated in the vertical direction by the 

 attraction of the axis of the disc ; but this elongation is very trifling if the 

 sphere present, as we have supposed, a diameter of 6 centimetres. 



10. The sphere of oil being thus suitably placed, we slowly turn the handle. 

 We then presently see the s\)lierejiaftcn at its poles and swell ovt at its equator, 

 and we thus realize on a small scale an effect which is admitted to have taken 

 place in the planets. 



However, although the results may be of the same nature in the case of the 

 great planetary masses and in that of our little masses of oil, I must not omit 

 to remark here that there is an essential difference between the forces which are 

 in play in the two cases. In the first, the force which tends to give to the great 

 planetary mass a spherical figure, and against which the centrifugal force acts, 

 is universal attraction ; in the second, the force which acts the same part with 

 regard to the small mass of oil, is molecular attraction, which is subject to 

 different laws. But as, on either hand, the aggregate of the actions reduces 

 itself to a contest between centrifugal force and another force tending to pre- 

 serve the spherical form of the liquid mass, it appears that the results must be 

 analogous, if not identical, with respect to the figure which that mass assumes. 



[This, we do not think, is quite correct. The forces which produce the equi- 

 librium of the ring are as follows : First. The centrifugal force which tends 

 to throAV the atoms from the centre of motion. Second. The force developed 

 by the external and internal horizontal curvatures, the direction of wliich is to- 

 wards the centre. Third. The force developed by the external and internal 

 vertical curvatures, one of which acts towards the centre, and the other from 

 the centre. The roundness of the ring is caused by the combined action of the 

 external and internal curvatures, which, under no circumstances of velocity of 

 rotation, would produce a flattened ring — J. IL] 



In order to observe, in all its beauty, the phenomenon on which we are en- 

 gaged, the handle must, at first, be turned with very little velocity — a tu/n in 

 five or six seconds. The effects are even then very decided. If we after- 

 wards apply a somewhat greater velocity — for example, a turn in four seconds — 

 tlie flattening at the axis, and the swelling at the equator, arc seen to be more 

 considerable, and they are further augmented by increasing the velocity of the 

 handle to one turn in three seconds. Before proceeding further we may remark 

 that, in these experiments, the handle must not be turned too long, for the masi3 

 of oil which, in the first moments, presents exactly a figure of revolution, 

 eventually loses this foi-m. At each fresh trial, therefore, the system must be 

 left to repose. The oil then resumes its spherical form, and slowly, of itself, 

 replaces itself in the proper position. The change of form which supervenes 

 when too many turns are given to the disc occasions results of a particular kind, 

 and which are not without interest. I shall speak of them by-and-by, (§ 22.) 



11. Now, if instead of moving the handle slowly a considerable velocity is 

 given to it, as two or three turns in a second, new, and very curious, phc- 



