PROCEEDINGS OF THE OHIO ACADEMY OF SCIENCE. 329 



graphs the sound-waves as they progress through the model. 

 Thus are shown in a highly instructive manner the original sound 

 and its behavior after reflection from walls, ceiling and fixtures 

 throughout the room. Since the size of the model bears to the 

 length of the wave from the spark a relation similar to that which 

 the size of the room bears to the average wave from the human 

 voice, the results are directly applicable, and have proved of 

 much value in designing audience-rooms. 



The other acoustic research is more widely known. It is 

 not always that a successful experimenter is also a finished 

 lecturer. Many in this room have doubtless heard Professor 

 Dayton C. Miller's admirable expositions of the photography and 

 analysis of sound vibrations. In this case, however, there was 

 a good deal of previous work. To speak only of some which 

 has been done in this country, as long ago as 1878 Professor 

 Blake of Brown University, one of the pioneering workers on 

 the telephone, devised and used an instrument for photographing 

 sound waves, consisting of a mirror attached to a telephone dia- 

 phragm, in a manner which while crude by the side of Dr. 

 Miller's refined dispositions, differs but little from them in prin- 

 ciple. Bevier of Rutgers College, enlarging the record of a 

 phonograph cylinder, as others also have done, studied several 

 of the vowel-sounds, analysing them by the application of Four- 

 ier's theorem. It may therefore be worth while to point out just 

 what Miller has done beyond devising, and using with rare ex- 

 perimental skill, apparatus more delicate and complete than that 

 of his predecessors. 



In the first place he has seen, as no one else has seen, that 

 the problem must be attacked in a large way. Most of the 

 earlier workers recoiled from the magnitude and the difficulties 

 of the task, making perhaps a few photographs, or analysing a 

 few curves, pointing out the way to possible results of value, 

 but attaining them to only a small degree. Miller has photo- 

 graphed thousands of waves, and has analysed hundreds of 

 curves, often to thirty partials. The mathematical analysis of 

 such a curve by any of the applications of Fourier's theorem is 

 a matter of hours or days. The mechanical analyser, devised 



