418 Prof. J. J. Thomson and Mr. H. F. Newall. [Dec. 17, 



Fig. 2. 



surface tension remains spherical as it descends. In fact, we may say 

 that, with some few exceptions to be noticed later, rings are formed 

 only when a liquid is dropped into one with which it can mix. This 

 is important, because surface tension has been supposed to play an 

 important part in the formation of these rings ; it is difficult, how- 

 ever, to see how any appreciable surface tension can exist between 

 liquids that can mix, and as far as our experiments go they tend to 

 show that it is only the absence of surface tension which is necessary 

 for their production. There are, as we shall show later, many cases 

 where rings are formed under circumstances in which there is no 

 possibility of capillary action, such as when the liquid into which the 

 drop falls is the same as the drop itself. As capillarity was found not 

 to be involved in the production of the rings, it seemed interesting to 

 investigate the subject further, and the following investigation was 

 undertaken with this object. As the result of our experiments we 

 have been led to a theory of the production of these rings of which 

 we shall now give a brief sketch in order to render the sequence of 

 our experiments more intelligible. 



Let us suppose that a spherical drop falls into a liquid ; the motion 

 of the liquid surrounding the drop will at first be much the same as 

 if a solid sphere of the same size were to fall into the liquid. Now, 

 when a sphere moves through a liquid the tangential velocity of the 

 liquid is different from the tangential velocity of the sphere, so that 

 the liquid flows past the sphere. If the sphere be fluid as well as the 

 medium in which it moves, there will not be an absolute discontinuity 

 in the motion, but only a very rapid change, so that there is a finite 

 alteration in an exceedingly small distance. This alteration is equiva- 

 lent to a vortex film covering the sphere, the lines of vortex motion 

 being horizontal circles, and if the liquid be viscous the vorticity in 

 the film will diffuse inwards and outwards. As the drop falls the 

 resistance makes it get flatter and flatter until it becomes disk-shaped ; 

 by this time, however, it is full of vortex motion, and as the disk- 

 shape is an unstable arrangement of vorticity, the disk must break up 

 into the stable arrangement — that of an anchor ring. This is a 



