554 



HANDBOOK OF PHYSIOLOGY 



NEUROPHYSIOLOGY 



FIG. 5. The ampulla and semicircular canal in the living 

 state before and during angular acceleration. The cupula, 

 situated on the top of the crista, traverses the entire lumen of 

 the ampulla. A, the cupula in its normal position; B, the cupula 

 during angular acceleration. Note the shift of position of the oil 

 droplet in the endolymph during acceleration. [From Dohlman 

 (24)-] 



endolymph flow in the horizontal semicircular canals 

 placed strictly in the horizontal plane, no flow is 

 thought to occur in the vertical canals (30, 61). 



Let us consider the case of rotation to the right 

 after adjustment of the head so that the two horizontal 

 semicircular canals, the synergic pair, are in the hori- 

 zontal plane. During such a rotation the left ampulla 

 is ' leading' its canal while the right ampulla is ' trail- 

 ing' behind its canal. The endolymph, owing to the 

 moment of inertia, will cause a deflection of the cupula 

 of the right horizontal canal in the ampuUopetal direc- 

 tion and an ampullofugal deviation of the cupula of 

 the left horizontal semicircular canal. When the rota- 

 tion is stopped, the two cupulae will be deflected in 

 the opposite direction owing to a backflow of endo- 

 lymph. Although movements of the endolymph stop 

 in about 3 sec, the cupula seems to take about 25 to 

 30 sec. to return to its resting po.sition. During this 

 time, the subject will experience a sensation of rota- 

 tion in the opposite direction. The stimulus to the 

 cristae obviously arises from the swinging of the cupula 

 set up by the endolymph. However, the cupula, being 

 elastic, returns to its original position if the speed of 

 angular stimulation becomes constant. Therefore, no 

 response to movements of steady velocity occurs be- 

 cause the endolymph, subject to the frictional influ- 

 ence of its enclosing walls, takes up the motion of its 

 canal and stimulation subsides. 



The threshold for perception of angular accelera- 

 tion has been studied by various methods (rotating 

 chair, torsion swing, after-sensation time) and is now 

 rather well-established. The torsion swing appears to 

 be the most sensitive method for measuring this 

 quantity (49). The minimum value for the perception 

 of rotation varies with the indicator used and with 



the method of computation. Mach (72) and Dodge 

 (22) found a th reshold value of 2 ° per sec,- The prod- 

 uct of the time and the acceleration required to reach 

 the threshold of rotational .sensation is constant. Thus, 

 for reaching the threshold, the required acceleration 

 is the greater, the shorter the time of its action. The 

 lowest values reported for the human threshold are 

 0.2° per sec- (i 13) and 0.5° per sec^ (5, 23, 49) 



Some experimental results have led to the conclu- 

 sion that the crista is a unidirectional receptor, capa- 

 ble of being stimulated only in one direction but 

 irresponsive to deflection in the opposite direction 

 (16, 70, 71, 78, 109, 123). Ewald (29) demonstrated 

 that an ampuUopetal cupular deviation in the hori- 

 zontal semicircular canals evokes a stronger reaction 

 than a corresponding ampullofugal deflection. In the 

 vertical canals the efl'ect of ampullofugal flow is more 

 marked. There is no explanation for this functional 

 diff'erence, between the horizontal and vertical canals, 

 a diff'erence emphasized repeatedly by many authors. 



More recent experiments, however, speak in favor 

 of a bidirectional function of the semicircular canals 

 (14, 61). The clearest evidence comes from experi- 

 ments with electrical recording of the action poten- 

 tials set up in the primary receptor fibers under condi- 

 tions of natural stimulation (68, 69). By dividing the 

 intracranial portions of the different nerve branches 

 from the labyrinth into very slender filaments, it has 

 been possible to obtain o.scillographic records of the 

 action potentials occurring in response to various 

 kinds of stimuli, and under favorable conditions it is 

 possible to continue the process of subdivision until 

 only one or two sensory units are in functional con- 

 nection with the recording device. Ashcroft & Hall- 

 pike (6) and Ross (94) made the first successful 

 attempts to exploit this possibility, using frogs. 

 Mowrer (80) recorded from the vestibular nerve of 

 the common painted terrapin. Later Lowenstein & 

 Sand (68, 69) made similar recording from the dog- 

 fish and ray and Ledoux (58) from the frog. They 

 demonstrated a clear bidirectional response against a 

 background of a resting discharge which is present 

 even when the animal is in a state of absolute rest; 

 their findings are in complete agreement with the 

 assumption that the cristae are stimulated as a result 

 of positive and negative angular acceleration. In- 

 creases or decreases in the resting discharge rate of 

 the sensory cells in the crista are brought about by 

 the deformation of their hair processes during deflec- 

 tions of the cupula. In the horizontal canals excitation 

 occurs, and an increased impulse discharge can be 

 recorded when the cupula is deflected in an ampul- 



■71; 



