216 ANNUAL REPOET SMITHSONIAN INSTITUTION, 1912. 



was shown to me by Prof. Wood. After thus taking a bubble between 

 my hands, I gently separate them until a neck is formed in the midst 

 of the bubble and then until this is divided into two. I now beat 

 the two bubbles together in a horizontal direction, and you observe 

 that they do not unite, but flatten each other and resist nearer 

 approach. If, on the contrary, they are made to approach in the 

 vertical direction, they unite mstantly and the approach is facilitated. 

 In the first case, the clean, smooth surface of the two bubbles did not 

 permit the air to be squeezed out, while in the second case the rough 

 surface of the liquid which drips away at the lower part of the upper 

 bubble broke the intervening layer of air. The two bubbles v/ere 

 then able to unite, either attached to a common face or septum, or, 

 if the latter broke before becoming visible, re-forming the original 

 single bubble. This operation may be repeated several times. 



What I shall now show you rests upon the foUowuig property 

 which bubbles possess: When their surfaces are smooth and clean, 

 they may be pressed against each other without there being actual 

 contact. If I place a bubble upon a horizontal ring just not large 

 enough to admit of its passmg through as a result of its own weight, 

 and if I push it downward with a liquid sheet spread over another ring 

 it will pass through at once. If v/e take care to remove all the drops 

 which may form upon the lower surfaoe, we shall be abJe to push it 

 upward agam. In no case does the liquid sheet really touch the 

 bubble. 



I blow a bubble under a ring and suspend from it another ring of 

 aluminum wire so as to be able to give it a shghtly elongated form. 

 Inserting the pipe, I blow into the interior another bubble, which I 

 detach. Tliis second bubble will descend into the interior of the first 

 until it finds a support at a certain parallel of latitude in its lower 

 hemisphere and will remain there as long as no drop suspended from 

 its lower part shall come into contact with the outer bubble. But 

 these drops can be removed mth the pipe. The outer bubble can 

 then be elongated and pulled downward by means of the aluminum 

 rmg so as to compress the inner bubble and give it the form of a 

 prolate spheroid, which can be swung around and around without 

 there being actual contact between the two bubbles during these 

 operations. 



I then remove the lower ring by peeling it off, so to speak, and then 

 withdraw the air present betv/een the two liquid sheets until it is 

 almost impossible to see between them. If I then blow in some 

 air again tangentially, the inner bubble wUl assume a rapid rotary 

 motion. The bubbles are too large to be easily projected upon the 

 screen. However, I can make these operations and other similar 

 ones more visible by simply utihzing the shadow thrown upon the 

 screen by the positive crater of an electric arc. 



