PHILOSOPHICAL SOCIETY OF WASHINGTON. 351 



In the summer of last year, IS11, with undiminished ardor, he 

 conliiiued his d^servations on sound; selecting this time Portland 

 harbor, Monhegan Island, and Whitehead light station, on the 

 coast of Maine. At the latter station, the abnormal phenomenon 

 of a region of inaudibility near the fog-signal, and extending out- 

 ward for two or three miles, (beyond which distance the signal is 

 again very distinctly heard,) had for several years been frequently 

 observed. This singular eifect is noticed only in the case of a 

 southerly wind when the vessel is approaching the signal from 

 the same quarter, and consequently with the wind adverse to the 

 direction of the sound beams, a condition of the wind which is 

 the usual accompaniment of a fog. The observation showed this 

 intermediate "belt of silence" to be well marked on board the 

 steamer both on ai)proaeliing the station and on receding from it 

 by retracing the same line of travel. Meanwhile the intermittent 

 signal whistle from the steamer was distinctly heard at the station 

 on both the outward and homeward trips of the vessel, through- 

 out its course. The next set of observations was made on the 

 opposite side of the small island, by directing the course of the 

 steamer northward ; and in this case the shore signal was dis- 

 tinctly heard throughout the trip, while the signal from the vessel 

 passed through tlie " belt of silence" to the observers at the sta- 

 tion. The hypothesis of a local sound shadow of definite extent, 

 is excluded by the simple fact that the regions traversed were 

 entirely unobstructed, the two points of observation — movable 

 and stationary — being constantly in view from each other when 

 not obscured by fog. The hypothesis of a stationary belt of 

 acoustic opacity is equally excluded by the uninterrupted trans- 

 mission of sound through the critical region in one direction ; 

 and this too whichever order of observation be selected. So that 

 in one of the cases the powerful whistle ten inches in diameter 

 blown by a steam pressure of 60 pounds, failed utterly to make 

 itself heard, while the sound from a much feebler whistle only six 

 inches in diameter and blown by a steam pressure of 25 pounds, 

 traversed with ease and fulness the very same space. The only 

 hypothesis left therefore is that of diacoustic refraction; by which 

 the sound beam from one origin is bent and lifted over the observer, 

 while from an opposite origin the refraction is in a reversed direc- 

 tion; and such a quality in tlie moving air is referable to no other 

 observed condition but that of its motion, that is to the influence 

 of the wind. Observations were afterward made at Monhegan 

 Island, on some of the more normal effects of the refraction of 

 sound by differences of wave velocity, all fully confirming the 

 supposition which had been so variously and critically subjected 

 to examination. 



The principal conclusions summed up in this last Report for 

 1817, are: 1st. The audibility of sound at a distance depends 



125 



