168 THE ROYAL SOCIETY OF CANADA 



station, have been plotted in curves a and c, respectively. The 

 ordinates of these curves have been increased proportionally so that 

 both curves pass through the point 10,000 on the ordinate for 860 

 feet. 



These curves indicate that the energy of the partials of low fre- 

 quency from 174 to 696 is absorbed so that it has lost more than half 

 of its value at a distance of two miles. The energy of the partials of 

 frequency from 1218 to 1740 is absorbed at a much greater rate, and 

 even if these partials contained as much initial energy as the lower 

 ones, yet the sound would be completely absorbed at a distance of less 

 than two miles. It will be worth while to compute the absorption co- 

 efficient when observations of greater weight are available. 



Three photographs of the sound from the diaphone were taken 

 from the deck of the boat while it was tied to the wharf, at a distance 

 of 960 feet from the horn, and on an azimuth about 30° to the west of 

 the acoustic axis of the trumpet. This station is on a line at right 

 angles to the axis througli the average station a b c oi Table II. The 

 average energy received at the wharf is represented by 1,378, which 

 is less than 14 per cent of that received on the axis directly opposite 

 where the energy is represented by 10,000. If the energy on the axis 

 at a distance of 860 feet is reduced to the distance 960 feet, it is repre- 

 sented by 8,025. The energy at the wharf at the same distance from 

 the horn but on an azimuth 30° to the west is 17 per cent of this. 

 Or, in other words, moving 30° in the arc of a circle from the axis of 

 the horn, reduces the energy received to one sixth of that received 

 on the axis. 



On the axis of the horn, the first five partial tones contain 74 

 per cent of the energy and the partials from six to ten contain 26 per 

 cent. At the wharf, the partials from six to ten contain only 3 per 

 cent of the total energy. This shows, what would be expected, that 

 the trumpet is more effective in directing the shorter wave-lengths, 

 or that the shorter wave-lengths do not deviate as far into the "acous- 

 tic shadow" as do the longer waves; it is an illustration of acoustic 

 diffraction. 



A series of photographs were taken of the sound from the 2-inch 

 experimental diaphone, when sounding under air pressures of 18-7 

 lbs., 23-5 lbs., and 28-8 lbs., per square inch, and with and without 

 the resonating trumpet. The phonodeik was placed at a distance 

 of about 100 feet from the diaphone, on a line at right angles to the 

 axis, through the end of the trumpet. This location of the observing 

 station was an unfavorable one, and though the curves have been 

 fully analyzed, the details need not be given. Two interesting facts 

 are indicated. 



