1 10 THE ROYAL SOCIETY OF CANADA 



tion of aerial plane waves involving this large degree of compression 

 indicates clearly that these sound-waves of large amplitude fall in the 

 régime of abnormal propagation which has been studied by several 

 writers. Considering the case of plane waves of large amplitude, it 

 has been shown that the wave changes its form as it advances through 

 the medium and ultimately tends to form a "bore" in which the 

 region of high compression and rarefaction travel very close together: 

 owing to adiabatic heating and cooling in the compressed and rarefied 

 portions of the wave, temperature equalization by thermal conduction 

 tends to take place across these two regions. As a result of this energy 

 dissipation the amplitude of compression is reduced until conditions 

 of normal propagation are reached and further energy losses are small. 



(iii). Although in practice divergence in three dimensions would 

 have a tendency to minimize the efïect referred to, and to establish 

 more promptly conditions of normal propagation, measurements of 

 the falling ofï of sound intensity indicate that severe energy losses 

 occur in the immediate neighbourhood of the siren throughout a dis- 

 tance varying between 1,000 and 2,000 feet. Beyond this distance 

 the sound waves are capable of travelling very long distances on a 

 calm day with very little energy loss. 



(iv). Measurements of sound intensity carried out over distances 

 of several miles on various occasions (15 days) gave no general law of 

 falling off of amplitude with distance owing to the heterogeneity of 

 the atmosphere as a medium for the transmission of sound. The 

 theoretical "inverse square" law for the falling off of the intensity is 

 utterly unreliable as a basis for calculation even on an apparently 

 calm day. By far the most important factor responsible for very 

 great variations in the intensity of sound is the effect of wind. Wind 

 blowing against the direction of sound propagation results in refraction 

 of the wave-front upwards and the formation of sound-shadows at 

 the surface of the sea. 



(v). The results of the acoustic surveys support the view (due 

 to Major G. I. Taylor) that the high degree of attenuation of sound in 

 a wind may be explained by the eddy-structure of the atmosphere 

 which recent meteorological research has brought to light. There is 

 reason to believe that extinction due to this cause is especially severe 

 within a half-mile radius of the siren. It is probable, also, that eddy- 

 motion in the immediate neighbourhood of the fog-signal apparatus 

 due to the presence of buildings in the vicinity may have a detrimental 

 effect in causing the intense sound-waves to dissipate their energy 

 content by the formation of vortices. 



