influence of Vibratory Motions. - 435 
‘The thread, we must remember*, generally divides. itself into 
three parts, of which the extreme ones proceed to reunite them- 
selves respectively with the two large masses between which the 
thread extended itself, whilst the intermediate part at the same 
time contracts symmetrically above and below by expanding itself 
horizontally, so as to form the spherule under consideration. In 
virtue of this simultaneity and symmetry of action, the small 
portion of liquid in question attains the spherical form towards 
which it tends; but attaining this form with an acquired velo- 
city, it necessarily goes past the same in such a manner that its 
vertical diameter becomes less, and its horizontal diameter greater, 
than that of the sphere of the same volume; hence the oscilla- 
tions of form in the spherules and the consequent ventral seg- 
ments and nodes. 
But the transformation of the large mass, suspended at the 
end of the thread and becoming detached by the rupture of the 
same, does not take place in exactly the same manner; in fact, 
an instant before this separation the mass in question was already 
set free at its lower part by the rupture of the thread between it 
and the preceding mass; here, therefore, the ruptures below and 
above the mass, and consequently the two contractions which 
tend to flatten the same in a vertical direction, are not simul- 
taneous. Moreover, as each of these contractions must be fol- 
lowed by an elongation, neither do the latter take place simulta- 
neously ; and the same remarks apply of course to the subsequent 
contractions and elongations. Hence each contraction of the 
bottom of the mass will take place wholly or partially during an 
elongation of the top, and vice versd; but the first tend to in- 
crease, the second to diminish the horizontal diameter of the 
mass, so that their effects upon this diameter will more or less 
destroy one another; and if there is no vibratory imfluence, 
which, by the increased velocity which it imparts to the trans- 
formation, carries the diameter in question beyond that of the 
sphere, and thus determines an excess of pressure at the equator 
of the mass, this diameter ought to vary but little, and conse- 
quently a system of ventral segments and nodes will not be ob- 
servable in the troubled part of the jet. We see then, that, even 
under the sole action of the forces of figure, the masses which 
detach themselves at the extremity of the continuous part are 
necessarily the seat. of oscillations of form; these oscillations, 
however, can only be well developed in a vertical direction. We 
committed a slight error, therefore, when in paragraph 69 of 
the’ Second Series we asserted that, after becoming detached, 
the masses immediately formed themselves into spheres. 
§ 19. Let us return for a moment to the spherules. Whena 
* Second Series, § 62, 
2F2 
