286 DR. LOUIS VESSOT KING ON THE PROPAGATION OF SOUND IN THE FREE 
Appendix III.—ON THE THERMODYNAMIC MEASUREMENT OF ACOUSTIC EFFICIENCY. 
(i.) General Procedure in Taking Observations. 
The construction of the open-wound resistance thermometers employed in measuring temperature 
differences is clearly shown in the Plate (iv.). That on the low-pressure side of the diaphone (So) was 
situated about 5 inches from the edge of the piston, while that on the high-pressure side (R) was situated 
in the large valve admitting the “ sounding air,” as shown diagrammatically in fig. 3. Electrical connection 
to the Wheatstone bridge was made through an automobile spark-plug inserted in the cover plate of the 
valve. At the beginning and end of a series of observations a test was made of the insulation of the wire 
grids. It was found that after a long series of observations the high-pressure thermometer became 
slightly damp if left in position over night, owing to the accumulation in the large valve, in which it was 
situated, of moisture brought over with the air from the storage tanks. It was thus necessary on 
commencing a test to thoroughly dry out the high-pressure thermometer; the insulation then remained 
satisfactory during the entire series of observations occupying about three hours. 
During the series of observations the gas-engine operating the compressors were kept running 
continuously until the highest pressure to be employed in the tests had been attained. During this 
interval the mercury manometer connected to the storage tanks was read at known intervals, so that the 
rate of pressure increase could be determined at each pressure. The small relay-valve connected to the 
large valves admitting the “ driving ” and “ sounding ” air to the diaphone was operated by one of the 
observers (H. H. H.) to give a 6-second blast as determined by an accurate stop-watch. The same 
observer also noted accurately the fall of the mercury manometer in this interval. At the same time the 
writer noted the deflection of the galvanometer of the Wheatstone bridge to which the differential 
thermometers were connected Each observation was repeated, the battery current of the Wheatstone 
bridge being reversed in the interval. The mean pressure fall in the 6-second interval is entered in 
the accompanying tables together with the difference of resistances of the differential thermometers 
corresponding to the observed galvanometer deflections. 
After a series of observations with the diaphone sounding has been taken, the diaphone piston was 
removed and three small pieces of rubber packing inserted between the down-stream face of the driving 
head and the opposite face of the cylinder. By means of a set-screw inserted in the centre of the cover- 
plate the piston as a whole was not only prevented from vibrating but could be moved forward slightly 
against the pressure of the rubber cushions until the opposing ports were in such a position that the 
pressure-fall of the manometer in 6 seconds was the same at a given pressure as when the diaphone was 
allowed to sound, indicating that the air consumption in the two cases was the same. This adjustment was 
made at about 20 lbs. pressure (atmospheric); an inspection of the series of pressures recorded in the 
accompanying tables shows that, except at some of the higher pressures, the two series agree fairly well 
throughout. Test No. 1 was carried out without any alteration of the existing adjustments. Test No. 2 
was carried out with the large valve admitting sounding air set so that the air consumption was roughly 
halved. Before commencing Test No. 2, the diaphone piston was removed and carefully cleaned and 
lubricated. The sounding-valve was then adjusted until the diaphone emitted its clearest and smoothest 
note as judged by the ear. A comparison of the data given in Tests 1 and 3 shows that at the same 
pressure the rate of air consumption is somewhat greater in the latter. The acoustic efficiency is 
considerably improved in Test 3. Test 4 was carried out with the rate of air consumption about halved. 
In Test 5 the supply of sounding-air was turned off' and the compressors stopped. The air from the 
storage tanks was then allowed to reciprocate the piston only. The fall of the pressure manometer in 
known intervals of time enables the rate of air consumption to be calculated. It will be noticed that only a 
comparatively small proportion of air is required to operate the diaphone piston. Unfortunately, 
temperature readings were not taken during this test. An upper limit to the rate at which energy is 
consumed in driving the piston may be made from the results of Test 4. 
