408 WRITINGS OP JOSEPH HENRY. [1856 



mains a certain, time on the sensorium, it is certain that no 

 sound is ever entirely instantaneous, or the result of a single 

 impression, particularly in enclosed spaces. The impulse is 

 not only communicated to the ear but to all bodies around, 

 which in turn become themselves centres of reflected im- 

 pulses. Every impulse must give rise to a forward and 

 afterward a backward motion of a small portion of the 

 medium. 



Sound from a single explosion in air equally elastic on 

 all sides tends to expand equally in every direction; but 

 when the impulse is given to the air in a single direction, 

 though an expansion takes place on all sides, yet it is much 

 more intense in the line of the impulse. For example, the 

 impulse of a single explosion, like that of the detonation of 

 a bubble of oxygen and hydrogen, is propagated equally in 

 all directions, while the discharge of a cannon, though heard 

 on every side, is much louder in the direction of the axis ; 

 so also a person speaking is heard much more distinctly 

 directly in front than at an equal distance behind. Many 

 experiments have been made on this point, and I may men- 

 tion those repeated in the open space in front of the Smith- 

 sonian Institution. In a circle 100 feet in diameter, the 

 speaker in the centre, and the hearer in succession at differ- 

 ent points of the circumference, the voice was heard most 

 distinctly directly in front, gradually less so on either side, 

 until in the rear it was scarcely audible. The ratio of dis- 

 tance for distinct hearing directly in front, on the sides, and 

 in the rear was about as 100, 75, and 30. These numbers 

 may serve to determine the form in which an audience 

 should be arranged in an open field in order that those on 

 the periphery of the space may all have a like favorable 

 opportunity of hearing, though such a disposition should 

 not be recommended as the form of an apartment where a 

 reflecting wall would be behind the speaker. 



The impulse producing sound requires time for its propa- 

 gation, and this depends upon the intensity of repulsion 

 between the atoms, and secondly, on the specific gravity of 

 the matter itself. If the medium were entirely rigid sound 



