NOTE ON THE IRON-TURNINGS CELL. 
5 
the slipping motion when the hubbies are very small, for such 
bubbles, enveloping points in this way, are never to be found. 
'On the other hand, bubbles of a size so considerable that they 
would at once escape by sliding, if attached in this manner to a 
point, may everywhere be seen clinging to the apparently smooth 
surface, of the bottom of a glass of soda water, as in fig. 2, 
and this is precisely the appearance they present at the orifice of 
..a narrow vertical tube (fig. 3). 
When such a bubble comes away from the tube it is because the 
surface becomes unstable and segments at the neck, and just as 
a hanging drop, when it . comes away, leaves a little adherent 
liquid, so a little air is left behind w'hen the bubble separates, 
and this will be sufficient to form the nucleus of a fresh bubble: 
hence we see a continual stream of bubbles coming away from 
the same place. 
But since in the case of boiling the greater part of the air is 
diffused each time through the vapour of the escaping bubble, 
the quantity which remains becomes less and less in a geo- 
metrical progression, which accounts for our finally getting rid 
of the air by long boiling, without the necessity of supposing any 
appreciable projoortion of it to be dissolved by the liquid. 
It therefore appears that the common notion that points and 
edges determine the formation of bubbles in boiling liquids and 
