STUDIES CORRELATING INCIDENCE OF MOTION SICKNESS 



411 



closely-spaced pulses to a highly damped 

 system, the time-integrated departure from 

 rest is small because each pulse rapidly 

 checks the effect of its predecessor; likewise, 

 when the pulses are spaced too far apart, 

 the same quantity is again small because it 

 is computed on a per unit time basis. For 

 intermediate spacings one obtains a maxi- 

 mum. It is interesting to note that when 

 this variation of T holding a and /3 constant 

 is carried out experimentally (2) the sickness 

 index exhibits a behavior like the one just 

 described for a. (b) Decreasing pulse dura- 

 tion (/3), holding amplitude and period 

 constant decreases a^ as the sine^; this is 

 also in qualitative agreement with the ob- 

 served behavior of the sickness index (17). 

 (c) Increasing a, holding /3 and T constant 

 should increase o-^ as a^ — unfortunately this 

 experiment was not performed by Wendt; 

 in support of the result we have only the 

 crude correlations such as were pointed out 

 above that sickness is more prevalent on 

 vessels subject to greater peak accelera- 

 tions. Wendt 's experiments with asym- 

 metrical waves (5) could lead to information 

 regarding the question as to whether the 

 receptors involved react similarly in both 

 directions, but the experiments cited here 

 do not permit any certain conclusion be- 

 cause the pulse durations were not all equal. 

 A word should now be said regarding 

 Wendt's conclusion that the vestibule is 

 too "rapid" a system to account for what 

 he calls the observed "natural frequency" 

 {T = 3.75 to 2.72 sec). The important 

 experiment for this decision is 2, Fig. 4. 

 The sickness indices for increasing periods 

 are 10, 46, 64, and 37. Taking the k^ and 

 2X values previously calculated (23) from 

 Lowenstein and Sand's data (30) for the 

 semicircular canal, however, we obtain cor- 

 responding 0-2 values proportional to 5.73, 

 7.42, 8.28, and 8.48— that is to say, the 

 period of maximum a probably would have 

 been even longer had Wendt used rays as 

 subjects. Thus, whereas we are in no posi- 



tion to insist that the labyrinthine organs 

 are the vibrators involved in this response, 

 we do not feel that these experiments rule 

 them out in any way. It also seems proper 

 to suggest that the terms "natural fre- 

 quency" and "resonance" be avoided except 

 when they are used in their conventional 

 physical meaning. If the labyrinthine re- 

 ceptors be overdamped vibrators, as has 

 been suggested in the previous section, then 

 they have no real resonance frequency or 

 natural period. Regarding the behavior of 

 0-^ as a function of period or frequency in 

 the experiment just discussed, it is evident 

 that there is a critical frequency, coo, at 

 which a^ or o- is a maximum, and should 

 these measures of stimulation prove to be 

 useful, it may be convenient to coin a word 

 for said frequency. For the moment, how- 

 ever, a more important incidental matter is 

 to note that for pulses of short duration 

 0}^ ~ K^. This follows from solving da^/do) 

 = 0, assuming that jS/2 is small enough so 

 that sin aj(/3/2) ^ (6/2)co. In principle, at 

 least, this relation provides for an indirect 

 experimental measurement of one of the 

 characteristics of the vibrator, k^. Having 

 measured this, the other characteristic, X, 

 could be obtained from the fact that the 

 smaller^^ frequency, coi/t, for which a^ is 

 (1/A;)th of Co, is related to coo by. 



X = 



Vk 



^ V oOQ — COl/A ; I 



— 1 |_ 2(ayk J 



We have said above that Wendt and 

 others have thought that the stimulating 

 power of an imposed wave resides in its time 

 characteristics as well as in its peak accelera- 

 tions. On structural and analytic grounds 

 this view seems amply justified, as we have 

 tried to show. It also receives indirect 

 support from considering the high accelera- 

 tions which some experimentalists build 

 into their devices in order to match the 



" There is, of course, also a larger frequency, 

 but this is not to be used in the equation for X. 



