20 



SE.YSITIVE FLAMES AND SOUND-SHADOWS. 



cesser of Prof. Tyndall in tlie chair of Natural Philosopliy at the 

 Royal Institution in London. 



Let the whistle be supplied with a continuous blast of air, or 

 any compressed gag, at steady pressure. Four or five feet away 

 from it is placed the nozzle of the burner from which the flame 

 issues. Its sensitiveness may be regulated at will by means of the 

 stop-cock and the water manometer gauge. Turning on the blast 

 through the whistle, the flame flares. Let the open hand be held 

 up between the two ; the flaring ceases. The nozzle of the burner 

 is in the acoustic shadow cast by the hand. If this result is not 

 successfully attained at the first trial, the sensitiveness of the 

 flame may be slightly modified to suit the conditions. The case 

 is entirely analogous to that of the glass bottles in the experi- 

 ments in San Francisco Bay. 



By using a small mirror to reflect the sound-waves, their 

 lengths may easily be measured in mid-air. Let the mirror be 

 put a few inches behind the flame and moved slowly toward this 

 or away from it. At certain distances the flame is observed to 

 flare violently, and at certain other points it becomes quiet, 

 though the sound has not been varied. Reflected waves are 

 meeting advancing waves. Where they meet in like phases, their 

 effect on the flame is intensified. But if the position of the mirror 

 is so adjusted that the flame is at a point where the opposing 



waves meet in unlike phase, these 

 neutralize each other and the flame 

 ceases to be agitated. The case is like 

 that of producing loops and nodes on 

 a string attached at one end to a vi- 

 brating body and fixed at the other 

 end. A series of sinusoidal curves 

 travel over its length, and are re- 

 flected from the fixed end, producing 

 the so-called stationary waves (Fig. 

 0). A returning sinusoid is super- 

 imposed on an advancing sinusoid, 

 producing two loops, with an inter- 

 mediate nodal point of rest and a 

 node at the end. The whole sinusoid 

 represents a wave-length, and the 

 distance from node to node a half 

 wave - length. The distance through 

 which the mirror is moved from one point of flame quiescence 

 to the next is a half wave-length for the pitch yielded by the 

 whistle. In some experiments thus conducted by the writer, 

 this distance was found to be a trifle over half an inch. The 

 whole wave-length was 1*05 inch. Assuming the velocity of 



Fig. 5. 



