350 FIGURES OF EQUILIBRIUM OF A LIQUID MASS 
arranging that the axis of revolution shall be horizontal ; the rings of 70 milli- 
metres, which are then vertical, should be previously placed at a distance of 
110 millimetres apart; each of them is attached, by its lower part, to a vertical 
iron wire, and the wires are themselves fixed, at their lower extremities, in a 
plane table of iron, which supports the whole system; finally, these wires are 
enveloped with cotton, that the oil may not adhere to them, (2d series, § 9.) 
A cylinder (Fig. 13) is first formed between the two rings, then we gradually 
diminish the volume of the mass by means of the small syringe. If, when the 
neck is not more than about 30 millimetres in diameter, we take care to remove 
the oil by only very small portions at a time, we shall succeed in reducing it 
to 27 millimetres, and thus obtain the result represented by Fig. 14. 
Now, it is evident that all these constricted figures with convex bases—figures 
which, like those considered in preceding paragraphs, may deviate as little from 
the cylinder as we choose—are still portions of unduloid, though taken differ- 
ently in the indefinite unduloid: while the middle of the one is occupied by 
the equator of a dilated portion, the middle of the others is occupied by the 
cercle de gorge of a constricted portion; the most extended of the former, 
except the transient unduloid mentioned above, is composed of an entire dilata- 
tion between two demi-constrictions, (Figs. 6 and 7,) and that represented by 
Fig. 14 is composed of an entire constriction between portions of two dilata- 
tions. 
> 
Lig 17 
§ 14. Resuming, now, our horizontal rings, with a view of placing, at will, 
the upper one nearer to or further from the other, let us again form a*cylinder 
between them, and, without changing their distance, gradually remove oil from 
the mass. If the ratio of the distance of the rings to their diameter is much 
less than in the last experiment of the preceding paragraph, the curvature of 
the bases, instead of augmenting in proportion as the constriction deepens, con- 
tinues, on the contrary, to diminish; and if this ratio does not exceed about %, 
the bases at length become absolutely plane. With a ratio still less, we may 
even proceed further; if the absorption of liquid is continued the bases become 
concave. Let us form, for instance, between our rings of 70 millimetres diame- 
ter, a cylinder 35 millimetres in height, (Fig. 17;) by gradual absorption of 
the oil, we shall see the bases sink more and more at the same time that the . 
constriction grows deeper, and, their curvature at length wholly vanishing, we 
shall have the result represented by Fig. 18. If we still continued to use the 
syringe, the bases would assume a concave curvature; but let us pause, for 
the moment, while they are yet plane. 
With such bases, the constriction comprised between the rings can no longer» 
(§ 7) pertain to the unduloid, and we arrive, consequently, at a new figure of 
revolution. Let us inquire what this is, in its complete state. We remember 
(2d serieg, § 4) that the pressure corresponding to an elementyag the superficial 
