•Mat 6, 1921] 



SCIENCE 



427 



pulled to the right exactly the opposite move- 

 ment of the eyes results. If as commonly be- 

 lieved, the stimulus is the pressure due to the 

 weight of the otolith, and 'if that pressure, 

 shifted to right by inclining the body to the 

 right acts by stimulating more strongly the 

 epithelium on the right hand side surely the 

 artificial application of pressure should pro- 

 duce the same result but just the reverse 

 actually happens. I believe that the actual 

 process is as follows: When the right side of 

 the otolith is pressed upon it is displaced to 

 the left just as one saucer standing in another 

 is displaced to the left when one pushes 

 straight down on its right hand margin. 

 This, then, is the same displacement as that 

 which occurs when the body is tilted to the 

 left. It is not the pressure but the displace- 

 ment due to the weight of the otolith which 

 brings about the normal stimulus. It is the 

 direction of the displacement which deter- 

 mines the direction of the resulting compen- 

 satory movement. I have ventured to sug- 

 gest that the displacement gives rise to differ- 

 ences of tension which act on the sensory 

 structures in a manner analogous to the 

 effects of different tensions on the vagus 

 endings in the lungs. It will be seen in the 

 following section that this wholly unexpected 

 result is in accord with the conclusions which 

 I had previously reached in regard to the 

 mode of stimulation of the ampullae. 



IV. THE MECHANISM OF THE PHYSIOLOGICAL 

 STIMULATION OF THE AMPULLiE 



An adequate discussion of the enormous 

 literature of this subject would far exceed the 

 limits of this paper. ITearly half a century 

 ago Mach, Breuer and Brown almost simul- 

 taneously published papers suggesting that 

 each canal functioned for the recognition of 

 rotational movements in its own plane. 

 Roughly stated the assumption was that 

 rotation of the head in the plane of any canal 

 would tend by the inertia of the fluid within 

 the canal to produce a current in the direction 

 opi>osite to that of the rotation. In the popu- 

 lar literature and in many of the text-books 

 a favorite statement of the theory is about as 



follows: Each rotation of the head in the 

 plane of any canal causes through the inertia 

 of the endolymph a current in the canal in a 

 direction counter to the rotation. The hair 

 cells of the ampullae stick out like paddles and 

 are deflected in the direction of the current. 

 The bending of the hair cells exerts pressure on 

 the nerve endings and produces the stimulus. 

 Mach, however, was too good a physicist to 

 believe that a current could be produced 

 under the conditions existing in the semi- 

 circular canals. He found that water placed 

 in a glass model of the dimensions of a human 

 semicircular canal showed no perceptible 

 current when rotating at a reasonable speed. 

 He affirmed on theoretical grounds that rota- 

 tion could cause a pressure but denied the 

 possibility of an actual current. Breuer 

 (1899) also found on anatomical grounds 

 that the theory in this form is not tenable, 

 for the hair cells do not project into the 

 endolymph, but are embedded in the gelatin- 

 ous mass of substance forming the cupula. 

 In more recent years Rossi (1914) has con- 

 structed a model of the dimensions of a 

 human canal and has been able to demonstrate 

 some movement of the contained liquid when 

 rotated; but the rate of rotation necessary 

 to produce visible movement is far beyond 

 the order of magnitude of the rate giving a 

 distinct physiological reaction. Moreover, 

 according to the very beautiful anatomical 

 work of A. A. Gray (1907-08) the semicircu- 

 lar canals in the squirrel and the rat have 

 each a diameter only 1/5 as great as in man. 

 These animals show no inferiority in labyrin- 

 thine reactions. In order to form a satis- 

 factory demonstration it would be necessary 

 to make the external diameter of the tube in 

 the model 0.25 mm. I have found (1912) 

 that in the horned lizard, Phrynosoma, rota- 

 tion at an angular velocity so slow as a 

 movement through 45° in 8 seconds caused a 

 distinct labyrinthine reflex. Under these cir- 

 cumstances a current is unthinkable. 



Direct experiment on the canals is more to 

 the i)oint than the foregoing theoretical con- 

 siderations. Loeb (1891) cut through or 

 excised portions of the canals in the dogfish 



