n of the mode in which centrifugal force acts in
overcoming molecular attraction, has been cited as a demonstration of
the truth of the nebular hypothesis. The conditions, however, are
entirely different. By means of clock-work he caused a globule of oil to
rotate in a mixture of alcohol and water _of the same density_, thus
entirely getting rid of the power of gravitation; and by increasing the
velocity he caused it to flatten out into a disc, and finally to project
a multitude of minute drops, which continued their revolutions so long
as the fluid in which they floated kept revolving by the motion of the
rotating spindle, _the divergent drops, the central mass, and the
surrounding fluid, being all the while of the same density_. But the
essential conditions of the nebular theory are, that _the central mass_
exert an attraction of gravitation upon all its parts, and _therefore be
denser than the surrounding ether or empty space_, and that _the cooling
and contracting rings be of a different density from the rest of the
mass_. Their divergence from the more fluid portion is supposed to arise
from their growing denser. And Reclus shows[204] that the divergent
drops owe their existence to the _expansion_, not to the _contraction_,
of the globule of oil. This experiment, then, contradicts the theory, so
far as it is applicable.

Plateau himself never adduced this experiment in support of the nebular
theory; but having, by way of illustration, spoken of the revolving
drops as satellites, and finding that expression misunderstood, he
corrected the error in a subsequent paper. He says: "It is clear that
this mode of formation is entirely foreign to La Place's cosmogonic
hypothesis; therefore we have no idea of deducing from this little
experiment, which only refers _to the effects of molecular attraction_,
and _not to those of gravitation_, any argument in favor of the
hypothesis in question; an hypothesis which _in other respects we do not
adopt_."[205]

_It was always contrary to the facts of astronomical science._ It has
accordingly been repudiated by the most eminent astronomers.

Sir John Herschel declares that the appearance of those groups, or
clusters, of stars, supposed to be formed by the condensation of nebulae
is quite different from that depicted by this theory, and that no traces
of the ring-making process is visible among them. He thus describes the
appearances of these groups; exactly the contrary of that de