TRANSACTIONS OF THE SECTIONS. 
15 
interruption of sound produced by a prolonged reverberation, and, 
comparatively, rarely from a deficiency in the voice of the speaker. 
The human voice had been heard distinctly at the distance of a 
mile and upwards in a calm atmosphere. Sir John Ross, Lieu¬ 
tenant Bowen, and many others, had borne testimony to this fact; 
and in the experiments above alluded to, in the open air, individuals 
conversed easily at a distance varying from 200 to 1000 feet, when 
it was calm. In many rooms, in consequence of the repeated re¬ 
flections between wall and wall, or between roof and floor, the sound 
of the voice might be heard continued many seconds after the indi¬ 
vidual trying the experiment had ceased to speak. In a newly 
fitted-up leaden chamber for the manufactory of oil of vitriol, the 
sound was heard prolonged for seven seconds; and when the dif¬ 
ferent notes of any chord were sounded successively by any indi¬ 
vidual, they were afterwards heard blending harmoniously in one 
compound tone. The leaden chamber was 80 feet long, 15 broad, 
and 16 high. In numerous public buildings similar effects are ob¬ 
served ; but if the walls be made rough and irregular, so as to lose 
all resilient power, and hung with drapery, the reverberation ceases. 
On the same principle, the reflecting power of the floor being taken 
away by a crowded audience, sound is very different in such an 
apartment from what is observed when it is comparatively empty. 
The distinction between the actual amount of sound and purity of 
intonation has not been sufficiently attended to. Much sound may 
be produced when the primary impulse is strengthened by combi¬ 
nation with the reflected sound of many preceding words, but it has 
none of that harmony and distinctness which is observed when the 
primary sound alone is allowed to fall upon the ear. The sound 
of cannon has been heard at the distance of 300 miles. Captain 
Stoddart’s account of the firing of cannon in the Baltic heard at this 
distance, affords the most ample and specific information on this 
subject. The sound of volcanic eruptions has been heard at a di¬ 
stance of nearly 900 miles. It cannot be doubted that the repeated 
reflection of preceding sounds must interfere most materially with 
those that succeed; and, from what has been above stated, it is ob¬ 
vious that such reflections must be continued frequently to a great 
extent in numerous apartments. 
In constructing buildings, the following circumstances require to 
be particularly noticed. 
In the most perfect form of building for the communication of 
sound, any reflected sound must be prevented from continuing so 
as to interrupt any new tone, by being thrown upon a non-reflect¬ 
ing floor. So long as the reflected sound comes up in time to 
strengthen the primary impulse before any new sound is heard, it is 
to be taken advantage of; beyond this it is injurious. A building 
having low walls, rough and irregular on the surface, an inclined 
roof terminating in a ridge in the centre, and having any elevation 
there that might be necessary, the material of which it is made 
having great reflecting power, with a floor matted and arranged so 
