WITHDRAWN FKOM THE ACTION OF GRAVITY. 413 



are too small to allow of easy verification ; in other words, the thread may be 

 said to constitute a slender tube, and the large spherule is a bubble whose ii> 

 terior is occupied by the alcoholic liquid. 



Let us remark, in connexion with this experiment, that nothing is better 

 adapted for the realization of that part of the nodoid (4th series, § 32) generated 

 by a node of the meridian line, when this node approximates to a circumference 

 of a circle (4th series, §35,) than the plane films which, after disunion, occupy the 

 two rings. If we pierce one of these plane films in its middle, the liquid ring 

 which the oil forms in retiring towards the metallic ring has, in relation to the 

 radius of the latter, but little thickness, (4th series, § 36,) and, as far as can be 

 judged by the eye, its meridian section is decide^ily circular : with a metallic ring, 

 forinstauce, whose radius is 35^^^, the width of the liquid ring, that is, the dis- 

 tance between its inner and outer circumference, is scarcely 3""". 



§ 5. The facility with which, under the conditions of my experiments, the oil 

 extends itself in thin films, connected with the fact above mentioned of the 

 formation of a laminar spherule, naturally led to the belief that large spherical 

 films of oil might be developed, or, in other words, that hollow bubbles of oil 

 of considerable size might be obtained in the alcoholic liquid by distending them 

 with the alcoholic liquid itself, as we obtain, in the air, bubbles of soap by in- 

 flating them with air. The experiment was completely successful. Analogy 

 readily suggested the process : in order to foi-m the bubble of oil a small mass 

 of this liquid was first made to adhere to the lower extremity of an iron tube 

 plunged vertically to a certain depth in the alcoholic liquid, and the liquid des- 

 tined to distend the bubble was then slowly poured through the other extremity 

 of the tube. 



§ 6. But this experiment, so simple in principle, exacts a number of precau- 

 tions which should be here pointed out. To facilitate the introduction of the 

 alcoholic liquid the tube should be funnel-shaped at its upper end, and, in order 

 that its position may be stable, it is necessary to adapt to the base of this ex- 

 pansion an iron disk of 7 to 8 centimetres in diameter, pierced in its. centre by 

 the tube and made to rest on the neck of the central aperture of the vesseL 

 The lower orifice of the tube should be furnished with a thin rim *of about 

 j^mm 5 j^^ width, the object of which is to prevent the small mass of oil designed 

 to form the bubble from partially rising along the exterior wall of the tube ; 

 the oil stops at the circumference of the little rim, conformably with the facts 

 described in § 13 of the 2d series, and arranges itself in a manner perfectly 

 symmetrical. Nor is the diameter of the tube a matter of indifference ; that 

 which has yielded me the best results is one of IG'"'". The section of the 

 FW. 4 system is represented by figure 4, 



It is evident that tfie alcoholic liquid with which the bubble 

 is to be filled should have precisely the same density with the 

 exterior liquid. This condition is satisfied without difficulty by 

 previously withdrawing, through the faucet at the lower part of 

 the vessel, a portion of the liquid contained in the latter, and using 

 this portion for the distension of the bubble. It is necessary, of 

 course, that the liquid should descend slowly and gradually into 

 the bubble, especially at the commencement ; it should, at first, 

 ^ ■ fall drop by drop, then in a slender thread, and should enter 



the funnel near its upper edge, in order that, gliding along the inclined wall be- 

 fore descending into the tube, it may have less velocity. But if, to accomplish 

 this operation, we are content to hold in the hand the flask containing the liquid 

 in question, we shall never succeed, whatever pains be taken, in giving to the 

 bubble the whole diameter which it might acquire, and this for two reasons : in 

 the first place, it is impossible to graduate, in a sufficiently regular manner, the 

 velocity of the flow, and the liquid sometimes arriving in too great abundance, 

 produces in the interior of the bubble such movements as cause it to burst ; la 



