WITHDRAWN FROM THE ACTION OF GRAVITY. 433 



meter, and it would be seen to progress in a continuous manner in proportion to 

 the ulterior attenuation of the film. In this case, the thickness which the film 

 had when the diminution of pressure commenced would be determined by the 

 tinge which the central space presented at that moment, and the half of that 

 thickness would be the value of the radius of sensible activity of the molecular 

 attraction. If, on the contrary, the pressure remains constant until the disap- 

 pearance of the bubble, we may infer from the tint of the central space the final 

 thickness of the film, and the half of this thickness will constitute at least a 

 limit, very little below which is to be found the radius in question. 



§ 31. I have made, in this view, a great number of experiments, of which I 

 proceed to give an account. At first a diameter of about four centimetres was 

 given to the bubble; it was then left to shrink to nearly two centimetres, and 

 sometimes even to one, before the wax ball was applied. In the first experiment, 

 the drop was next removed and the adjutage with the bubble introduced into 

 the interior of a small jar, the orifice of which was simply closed with a disk 

 of pasteboard ; the contact of the horizontal thread of the e3'e-glas3 of the 

 cathetometer with the summit of the image of the surface of the water in one 

 of the branches of the manometer was now established, and as equilibrium did 

 not immediately take place (§ 27,) the contact was readjusted from time to time 

 until it became stationary. Eight bubbles were observed under these circum- 

 stances until their disappearance : of these, seven burst before having passed 

 the first colors of the second order ; one only seemed to have attained the in- 

 digo of that order, but there was some uncertainty as regards this ; the great- 

 est duration was fourteen hours. 



As to the contact of the thread of the eye-glass with the image of the surface 

 of the water, it never, but in one instance, varied in the direction of a diminu- 

 tion of pressure, but, singularly enough, it sometimes varied by small quanti- 

 ties in the opposite direction. For one of the bubbles it Avas ascertained, by 

 measurements taken before and after these variations, that the pressure had 

 really somewhat increased. When such a variation was produced, it was with 

 a certain rapidity, and the manometer afterwards remained stationary, either 

 until the disappearance of the bubble or until a new variation in the same di- 

 rection. These variations are not owing to changes in the temperature, for 

 that of the apartment was very constant, nor do they proceed from an imper- 

 fect application of tho wax, for in that case the augmentation of pressure would 

 be continuous and accelerated. 



§ 32. These experiments alone might have furnished me with a result ; but 

 I was desirous of knowing why the colors of the bubbles proceeded no further. 

 Led to suspect that a slight chemical action between the iron of the adjutage 

 and the liquid a little altered the constitution of the latter in the vicinity of the 

 orifice, I fitted to this orifice, with sealingwax, a piece of glass tttbe of the 

 same exterior diameter and with walls suitably thin, and inflated at the free ex- 

 tremity of the tube a bubble which was introduced, as before, into the small 

 jar. Then, indeed, things took place after a different and rather curious manner : 

 the colors proceeded at first even into the third order, after which they retro- 

 graded gradually to the red and bluish green of the last orders, then grew pale, 

 and the bubble finally again became white as at the moment t)f its formation. 

 The thickness of the film therefore had been first diminishing, and afterwards 

 augmenting. The bubble lasted twenty-four hours. 



This phenomenon would have seemed inexplicable, had not an experiment 

 which I had made before equipping the adjutage with the piece of glass tube 

 furnished me the key to its solution. For the experiment referred to, a little 

 water had been poured into the jar, and the interior walls of this had been also 

 moistened : now a bubble placed in this atmosphere saturated with watery va- 

 por had in like manner lasted for twenty-four houi'S, and had burst without 

 having emerged from the red and gi-eeu of the last orders ; it had therefore 

 2S s 



