MECHANICS OF THE INNER EAR 69 



with twenty-two. The present writer at least is inclined to 

 doubt this. He believes that the theory, representing only an 

 approximation to what actually happens in the organ of hear- 

 ing, exaggerates the degree of this loss of intensity on the 

 part of the higher tone. He is also inclined to believe that 

 the theory exaggerates the relative intensity of the difference 

 tone 1, which was found to be six. In reality, this tone seems 

 to be somewhat weaker than is indicated by this number. 



Let us remember, now, the provisional assumptions 

 which we made in order to render the graphic representation 



of the movement of the partition as sim- 

 The third and P'^ ^^ possible. We may raise this ques- 



fourth provisional tion : Is not, perhaps, the above disagree- 

 assumptions ment between theory and experimental ob- 



recalled servation a result of one or more of these 



provisional assumptions? I shall demon- 

 strate that this is indeed the case. Or, more exactly, I shall 

 demonstrate that, if we omit one of these assumptions and 

 take into account in its stead the actual anatomical conditions 

 so far as these are known, we change the results of the theory 

 in such a direction as to diminish the exaggerated loss of 

 intensity of the higher primary tone and also the exaggerated 

 intensity of the difference tone. 



The partition was provisionally assumed to be of equal 

 width all along the tube. As a matter of fact, its width near 



the windows is only one-twelfth or one- 

 The partition is tenth (measurements differ somewhat) of 

 narrovsrer near what it is at the far end of the tube. And 



the windows further, it is to be noted that the width 



of the partition does not increase uniformly 

 along the tube, like the area between the dotted lines of figure 

 24, but that it increases first rather rapidly, later morfe slowly, 

 like the area between the curved lines. The figure, however, 

 does not represent the true relation between the width and 



