410 Dr. Gregory on the 



the direct and the reflected transmission of the sound from a 

 musquet fired by my side, and returned in an echo from the 

 front of the said storehouse. The distance from my station 

 to the front of the storehouse, determined carefully by a tri- 

 gonometrical operation, was 1523 feet. 



Of eight rounds fired from the musquet, I failed twice in 

 the appreciation of the interval between the sound and the 

 returning echo, from a very wrong estimate of its probable 

 duration ; and that from an erroneous impression as to the 

 time observed by Dr. Derham in a similar experiment*. Of 

 the remaining six rounds, the musquet pointed across the 

 river, the intervals were 2"-7, 2'"75, 2"-74, 2"'72, 2"-75, 2'"74; 

 their mean 2 7, 73' 



Next, three rounds were fired, the musquet being pointed 

 directly from the river; the intervals were 2"*7, 2"*73, 2"*76; 

 mean as before. 



Lastly, four rounds were fired along the bank, at an elevation 

 of about 45°; the intervals were 2"-75, 2"*7, 2"-73, 2"'74; 

 mean as before. 



Distance occupied by the direct and the reflected sounds 

 3046 feet. 



= 1116 feet velocity of sound across a surface of wa- 

 ter, half direct, half reflected ; therm. 66°. 



The near agreement of this with the former result on the 

 same day, serves to confirm the opinion that direct and re- 

 flected sounds move with the same velocity. 



Thursday, August 21, three o'clock P.M., barom. 29*86 

 inches, Fahr. therm. 64°; clear sunshine; wind scarcely per- 

 ceptible, westerly. 



Mortars were firing from the battery, and I took a station 

 3900 feet south of it. I observed the intervals between the 

 flash and the report in six successive rounds : they were 3' ,# 5, 

 3''-5, 3"-48, 3" '52, 3"-5, 3"'5, respectively: the mean being 

 3"-5. 



3 . 5 = -7=- = 11 14>f feet, velocity of sound, therm. 64°. 

 These are all the experiments in reference to the velocity 



* lie made it 3 seconds, by means of a half-second pendulum. My 

 erroneous recollection of his experiment led me to anticipate an interval 

 of between 4 and 5 seconds. I could not account for the supposed dis- 

 crepance until after my return home, when, on examining Derham's paper, 

 and computing the leal breadth of the river from my trigonometrical ope- 

 ration, I found the correspondence of the two experiments to be quite as 

 great as could be expected, considering the different natures of the chrono- 

 meters emploved, and the varying breadth of the river. 



of 



