262 W. R. HEBBLE WHITE. 



plates with varying apertures on the top of the float. For 

 quantities above 1*4 cubic feet per minute, this plate is 

 replaced by one having a hole sufficiently large to pass a 

 large proportion of this amount of air. The float now falls 

 considerably, but further increase in the rate of flow will 

 again cause it to rise and begin a new range. There are 

 three such " ranges" of readings, giving a total capacity of 

 S cubic feet per minute. The capacity might be increased 

 by enlarging the discharge holes, but this would decrease 

 the sensitiveness. For a larger plant, however, it would 

 be advisable to do so, 



4. Calibration. -Each range was calibrated separately, 

 and the curves are shown in Fig. 3. Calibration was 

 effected by the use of standard meter calibrating tanks. 

 The varying pressures of the air were read from a mano- 

 meter, and the necessary allowance made. 



5. Diagrams.— Sample diagrams taken by the instrument 

 are shown in Fig. 4, superimposed, with a common zero 

 line. The amount of air passing was varied by the air cock 

 on the exhaust main. It will be observed that in the 

 second and third ranges, the vibrations diminish very con- 

 siderably as compared with the first range diagram. 



6. Impulse Absorber.— The absorber apparatus, as at first 

 arranged, was an attempt to provide a reservoir in the air 

 passage, which, while free to take up, by change of volume, 

 the irregularities in the air flow, maintained at the same 

 time the constant pressure corresponding to that in the 

 instrument for the given rate of flow. This was easily 

 arranged by the use of a small gasometer in the air path,, 

 the float of which was balanced by a counter-weight 

 sufficiently large to give the required air pressure. The 

 gasometer rose and fell under the varying irregularities, 

 but the air passed on at a constant pressure to the recorder 

 and gave a good diagram. The gear was not automatic 



