Theory of the Pneumatic Paradox. 297 



being of a higher temperature than the surrounding air. This 

 explanation is set aside by the fact, that if a pair of bellows be 

 used instead of the lungs, the experiment succeeds equally well. 

 Another explanation published in his youth by Charles G. Page, 

 M. D., a gentleman well known to scientific readers, both in this 

 country and in Europe, for his inventions and discoveries in elec- 

 tro-magnetism and magneto-electricity, originally communicated 

 to this Journal, ascribes the phenomenon to currents of air which 

 strike against the movable disk nearly at right angles to its plane, 

 and which give it a tendency to adhere to the fixed disk. He 

 demonstrated the existence of these currents, by admitting into a 

 darkened room through a hole in the shutter, a beam of light, 

 and scattering a little dust in the beam, so that the direction of 

 any currents of air might be indicated by the motions of the par- 

 ticles of dust. These currents are caused by the air which issues 

 from between the disks, carrying with it some of the contiguous 

 air, into the place of which they rush nearly at right angles with 

 the disk. They are evidently inadequate to cause the adhesion 

 of the disks, since they cannot have a momentum greater than 

 that of the aforesaid contiguous air, which is much inferior to 

 that of the original blast. That they are not essential to the 

 adhesion of the disks, may be demonstrated by the following ex- 

 periment. Make of thin letter paper a hollow cylinder, of the 

 same diameter as the disks, eight or ten inches long, and tight at 

 both ends. Upon the mouth of a jar place a cover having a cir- 

 cular hole large enough to admit the cylinder without friction, 

 and through this hole sink the cylinder till it projects but little 

 above the cover. On applying the fixed disk to the superior base 

 of the cylinder, and blowing with a strong and long continued 

 blast through the tube, the cylinder may, against its own gravity, 

 be raised from the bottom of the jar, and even be lifted entirely 

 out of it. In this case it is obvious that the cylinder cannot be 

 sustained by currents underneath, and, of course, that such cur- 

 rents cannot be essential to the adhesion of the disks. The same 

 thing may be hkewise shown in the following manner. Let a 

 newspaper be pasted to a table by the edges in such a manner 

 as to occasion no tension, and, on applying the fixed disk to the 

 middle of it, and blowing strongly through the tube, that part of 

 the newspaper may be sensibly raised from the table. In this 



