H. V. Gill — Theory of Singing Flames. 



183 



the jet, the following experiment is of value. The 





gas near 



flames, as in fig. 4, were arranged so as to be as near 



ble to each other, the point of the mano- 



metric flame just reaching the base of 



the other. The axis of the rotating 



mirror was in the same plane as these 



two flames. On viewing the two images 



in the mirror, an eye placed in the 



plane of the flames sees the images as in 



fig. 5. 



The top row represents the image of 

 the singing flame. It will be seen that 

 what we had anticipated actually takes 

 place. The appearance of the flame also 

 supports our theory. When gas issues 

 under pressure the flame drags in the air 

 which surrounds it, thus presenting the 

 aspect of the flame of a Bunsen burner. 



The following facts show the impor- 

 tance of the pressure ot the gas in the 

 phenomenon. We have seen that if the 

 supply pipe be plugged near the jet that 

 the flame will not sing. This is clearly 

 between the two pressures is interfered 



as possi- 



marks that, with a tube 15 to 20 cm long, 



because the reaction 

 with. Tyndall re- 

 he was able to obtain, 



by varying the size of the flame and its position in the tube, 





a series of notes represented by the numbers 1, 2, 3, i, 5. He 

 says also that this experiment shows why it happened that 

 various experimentalists, who did not change the position and 

 size of the flame, had difficulty in obtaining desired notes in 

 their public lectures. We see clearly from our theory why 



