286 Mr. T. C. Porter on a Method of 
first, and lit at the jets (a) and (0), and these, if of giass, should 
be allowed a little time to grow warm before the pinch-cock 
admitting air from the gas-bag to the jets is unscrewed : 
this should be done very gradually, and the gas and air which 
issues at jet (a) adjusted until the issuing flame is quite blue. 
Fig. 2 (a). Fig. 2 (0). 
Qe 
| | 
/ 
E 
| 
The jet c should then be lighted and turned down very low. 
Jet b is then brought into the position in which it gives the 
loudest sound, and the final adjustment of the gas and air 
passing through (b) made by very gradually opening the 
pinch-cock which regulates the supply of air from the bag : 
if the reinforcement of the sound is still unsatisfactory, the 
supply of gas to (b) should be gradually shut off. In some 
instances the best results are obtained when almost pure air 
is issuing from 0. If the jets hiss and roar, as they are at 
times wont to do, ¢ may be turned up a little higher, or the 
position of jet 6 may be very carefully shifted further into, 
or away from the flame from jet a, or moved slightly from 
or towards jet a: in practice the adjustments are not difficult, 
and when all is right the reinforcement of the tones of the 
phonograph is very striking, being quite equivalent to the 
use of a small trumpet, and easily heard all over a fairly 
large room. Fig. 3 (Pl. XVI.) is the stereoscopic photograph 
of the flame, °37 times life-size, when it is in a condition to 
“play” well, and fig. 4 is a photograph of the general arrange- 
ment of the apparatus taken whilst the flame was “ playing.” 
It will naturally be supposed that if pure oxygen is used 
instead of air the results will be better: I have made the 
experiment, with the expected result, but the difference is 
not so great as one would expect, and in practice the manipu- 
lation of the jets is made so much more difficult, to say 
nothing of the increased cost, that 1 do not recommend the 
experiment. Hthylene and acetylene in place of coal-gas L 
have not tried, but there is no reason for supposing that they 
would vield much better results than those already obtained. 
jeé b 
4 
/ 

