298 Theory of the Pneumatic Paradoz. 
and in the preceding experiment, the tube should be at least a 
quarter of an inch in diameter. 
The most recent explanation of the pneumatic paradox that has 
come to my knowledge, is extracted in the number of the An- 
nals of Electricity, &c., for August, 1838, from a work on “ The 
Causes of Planetary Motions,” by Jabez Allies, Esq., who infers 
from the adhesion of the disks, the general principle, that “air ina 
state of agitation, or currents, whether it be cold, as from bellows, 
or warm, as from the mouth, is in a more rarefied state than the 
surrounding atmosphere.” This is certainly a very summary 
method of solving the difficulty. So far is this principle from 
being true, that it may, I think, be satisfactorily shown, that when 
air is expelled from the lungs or from a pair of bellows, it is in a 
denser state than the surrounding atmosphere. ‘The reason is 
obvious; the air, both in the bellows and in the mouth, is con- 
densed at the moment of its expulsion by mechanical pressure, 
and, as it issues into the open air, it meets with resistance to its 
expansion ; so that not only the current itself must be in a denser 
state than the surrounding atmosphere, but it must condense ina 
greater or less degree the air against which it is impelled. These 
reasonings are verified by the following experiment. Make upon 
a small, square rod, a tube of tissue paper, which, when the rod 
is withdrawn, will retain its form. Fasten this to a tube of me- 
tal, or other hard substance, of at least equal bore, and blow 
strongly through it, either with the bellows or the lungs. The 
sides of the paper tube will become convex, and, of course, the 
tube itself nearly cylindrical. This proves conclusively, that the 
current of air in the tube is more dense than the exterior air, since 
the tube, instead of being compressed, as would be the case, if 
the contained air were in a rarefied state, gains, according to the 
doctrines of isoperimetrical geometry, an actual increase of capa 
city by its change of form. Instead of a square tube, a eylin- 
drical one may be used. In this case, any rarefaction of the 
contained air, would be indicated by a greater or less degree of 
compression of the tube, which does not take place. The idea 
Most naturally suggested on observing the adhesion of the disks, 
and the correct one, is, that it must be caused by the rarefaction 
the interposed air; but the difficulty of assigning any satisfac- 
tory reason for the rarefaction, seems to have driven all who have 
