February 28, 1901] 



NATURE 



419 



graphic plate was set on edge in the dark room in such a 

 position that it would be ifluminated by the divergent beam 

 coming from the image of the arc when the shutter was opened. 

 A ruby lamp was placed in front of the plate, and rings were 

 then blown from a tube in front of the sensitive film. As soon 

 as a good ring, symmetrical in form and not moving too fast, 

 was seen to be in front of the plate, a string leading to the 

 shutter was pulled and the plate illuminated with a dazzling 

 flash. The ring casts a perfectly sharp shadow owing to the 

 small size and distance of the source of light ; the resulting 

 picture is reproduced in Fig. 2. The ring is seen to consist of 

 a layer of smoke and a layer of transparent air, wound up in a 

 spiral of a dozen or more complete turns. 



The angular velocity of rotation appears to increase as the 

 core of the ring is approached, the inner portions being screened 

 from friction, if we may use the term, by the rotating layers 

 surrounding them. This can be very nicely shown by differenti- 



FlG. 2. 



ating the core, forming an air ring with a smoke core. If we 

 make a small vortex box with a hole, say 2 cms. in diameter, 

 fill it with smoke and push very gently against the diaphragm, 

 a fat ring emerges which rotates in a very lazy fashion, to all 

 appearances. If, however, we clear the air of smoke, pour in a 

 few drops of ammonia and brush a little strong HCl around the 

 loweir pirt of the aperture, the smoke forms in a thin layer 

 around the under side of the hole. Giving the same gentle push 

 on the diaphragm, we find that the smoke goes to the core, the 

 rest of the ring being invisible, the visible part of the vortex 



Fig. 3. 



spinning with a surprisingly hi^h velocity. Considerable knack 

 is required to form these thin crescent-like vortices, the best 

 results being usually attained after quite a number of attempts 

 have been made. A drawing of one of these smoke-cores is 

 shown in Fig. 3. The actual size of the vortex being indicated 

 by dotted lines, it is instructive as showing that the air which 

 grazes the edge of the aperture goes to the core of the ring. 

 The experiment does not work very well on a large scalei, 

 though 1 have had some success by volatilising sal ammoniac 

 around the upper edge of the aperture by means of a zig-zag 

 iron wire heated by a current. 



By taking proper precautions we can locate the smoke else- 

 where, forming a perfect half-ring, as is shown in Fig. 4, illus- 

 trating in a striking manner that the existence of the ring 

 depends in no way on the presence of the smoke. The best way 

 to form these half-rings is to breathe smoke very gently into a 



NO. 1635. VOL. 63] 



paper tube, allowing it to flow along the bottom, until the end 

 is reached, when a ring is expelled by a gentle puff'. A large 

 test tube with a hole blown in the bottom is perhaps preferable, 

 since the condition of things inside can be watched. It is easy 

 enough to get a ring with most of the smoke in the lower half, 

 but to get a ring, one half of which is wholly invisible, the 



Fig. 4. 



smoke ending abruptly at a sharply defined edge, as shown in 

 the illustration, requires a good deal of practice. I have tried 

 fully half-a-dozen different schemes for getting these half-rings 

 on a large scale, but no one of them gave results worth 

 mentioning. The hot wire with the sal ammoniac seemed to be 

 the most promising method, but I was unable to get the sharp 

 cut edge which is the most striking feature of the small rings 

 blown from a tube. 



In accounting for the formation of vortex rings, the rotary 

 motion is often ascribed to friction between the issuing air-jet 

 and the edge of the aperture. It is, however, friction with the 



Fig. 5. 



exterior air that is for the most part responsible for the vortices. 

 To illustrate this point I have devised a vortex box in which 

 friction with the edge of the aperture is eliminated, or rather 

 compensated, by making it equal over the entire cross-section 

 of the issuing jet. 



The bottom of a cylindrical tin box is drilled with some 200 

 small holes, each about 1 7 mm. in diameter. If the box be 

 filled with smoke and a sharp puff of air delivered at the open 

 end, a beautiful vortex ring will be thrown off from the cullender 

 surface (Fig. 5). We may even cover the end of a paper tub 



QOQQ 



Fig. 6. 



with a piece of linen cloth, tightly stretched, and blow smoke 

 rings with it. 



In experimenting with a box provided with two circular 

 apertures I have observed the fusion of two rings moving side 

 by side into a single large ring. If the rings have a high velocity 

 of rotation they will bounce apart, but if they are sluggish they 



