14 W. B. Rogers on Sonorous Flames. 
It thus appears that the vibratory movements of singing 
flame are directly traceable to mechanical conditions which have 
been actually proved to exist in air-jets at the moment of their 
efflux. We are therefore not called upon to resort to conjectural 
properties of gaseous streams in order to explain them. 
As to the mode in which these primary vibrations are exci- 
ted at the aperture, whether by friction or some other special 
cause, we are not perhaps prepared in the existing state of our 
knowledge to form a satisfactory theory. But whatever may be 
their cause, few I think will be disposed to explain them by 
any assumed cohesive action of the particles of the gaseous ~ : 
stream. The analogy between jets of air and those of liquids so 
far as these vibrations are concerned, has been shown by Masson 
and Sondhauss to be only partial, but even were it more com- 
Le we could have no difficulty in admitting that the impelling ] 
orces, in the one case the elasticity of a pent-up mass of air, in — ; 
the other the weight of a liquid column, must produce effects in 
many respects alike upon the issuing streams of gas and liquid. 
In me 25 2 to apply the theory of liquid jets so beautifully 
developed by Plateau to jets of gas, we are obliged with Prof. 
so long as this state continues can we conceive the pict to 
strong compression the gas is brought near its point of 
liquefaction, we may conceive this 
ulsion to grow rapidly less and thus to cause the departure 
an.efficient energy of cohesion among the particles, the 
lesemg 
to form 
Pea 
1) 
orTring 
eet 
Thi 
ies of shape between sheets of flame and 
