Chemical and Physical Papers. 91 



ably hundreds of times more intense than those with which we 

 ordinarily have to do. When working with light the suspended 

 parts could be put within an air-tight vessel and so protected from 

 -air currents, but in the case of sound measurements this is im- 

 possible, as that would shut off the pressure that we desire to 

 measure. This puts a very definite limit to the delicacy of the 

 suspended system that may be used in such an instrument. Some 

 form of microphone, properly calibrated, will perhaps prove in the 

 ■end the most satisfactory means of estimating sound intensities. 



