A SELECTIVE HOT-WIRE MICROPHONE. 
421 
of free convection from the grid, but is more nearly the component of the free convection 
along the axis of the microphone. That is, V 0 is approximately equal to V cos 0, 
\ being the velocity of free convection from the grid. If this were strictly true, we 
should have V 0 = 0 when 6 = ^, but this is not generally the case, probably owing 
to the fact that the loops are slightly displaced and do not lie in a plane which is just 
at right angles to the microphone axis, and it may even happen that they are not all 
in the same plane. For example, in the case of the grid used in the experiment just 
described, it was found that 
and 
V 0 = 1 *225 cms. per second when 0=0, 
V 0 = 0*25 cm. per second when 6 = — ■ 
A 
As an approximation we may take 
V = 1*225 and V 0 = Y cos(0 + a) when 0 is 
near —, so that 
2 
Whence 
1*225 cos 
= 0*25. 
and we should have V 0 = 0 when 6 = 101°. 
When V (J = 0 the expressions for the resistance changes produced by an oscillatory 
air-current (§ 6) reduce to 
£Rj — |-aU 2 , 
= 0 , 
£R 3 = — |-«U 2 cos 2 pt ,' 
so that the total resistance change is made up of two parts only—a steady change and 
the octave—while the fundamental vibration is completely suppressed. Now the 
curve in fig. 15 shows that in that particular experiment the fundamental was suppressed 
when 6 was about 125 degrees, and this is indeed about the usual value of 6 for this 
phenomenon to occur, while (as in the above example) V 0 vanishes when 6 is at most 
about 100 degrees. It appears therefore, that merely writing U sin ft instead of U, in 
the equation connecting U and <fR for small steady velocities, does not in this case 
explain all the observed phenomena. 
A satisfactory explanation of this difficulty has not yet been found, but the hypothesis 
that a jet may be formed in the neck of the resonator may be put forward. The 
possibility of this occurring in the mouths of ordinary resonators is discussed by 
Rayleigh (“ Theory of Sound,” vol. II., § 322). If a jet were formed in the neck of 
the resonator, then, in order to account for the observed phenomenon, it would have to 
3 M 
VOL. CCXXI.-A. 
