March 17, 1899.] 



SCIENCE. 



399 



VII. Preceding Case (IV.) Adjusted for Ro- 

 tary Polarization. 24. — If a special axle be 

 provided with the cams alteraately in op- 

 posite phase to the normal occurrence, but 

 otherwise equal in amplitude and wave- 

 length, and if the corresponding balls be 

 painted red and white, the two circu- 

 larlj' polarized waves occur simultaneously. 

 Similarly, the two plane polarizations at 45° 

 and at 135° occur simultaneously; etc. The 

 former case is interesting in relation to ro- 

 tary polarization, as will be more fully indi- 

 cated below ; for the two circular motions 

 may be compounded by the device shown 

 in Fig. 10, and a harmonic curve plane po- 

 larized in the vertical or the corresponding 

 wave will result (cf. §40 et seq.). 



To obtain rotation of the plane of polari- 

 zation by this method the alternate cams 

 on both the front and rear axle would have 

 to be set for some other wave-length in the 

 manner stated. 



(vertical) axles of which are at the angles 

 of the cranks, as far apart as the cams, 

 and all arranged along a straight line paral- 

 lel to the cam axle. The short shanks of 

 the bell cranks now carry a series of -J" 

 balls, which, under present conditions, must, 

 therefore, vibrate nearly pai'allel to the cam 

 axles, i. e., longitudinally right and left in 

 the line of advance of the wave, whereas 

 the thrust of the levers* is harmonically to 

 and fro. 



In practice the long shanks are open sec- 

 tors of wire, swung so as to clear each oth- 

 er's axles. 



In this way the alternate compression 

 and rarefactions of such a wave are re- 

 markably well shown (cf. Fig. 11), the 

 sinuosity in the line of particles being 

 negligible at least to the observer in front. 

 The balls approach each other to about 

 5/8" between centers (all but contact in 

 the compressional phases) , while they sepa- 



FlG. 11. Adjustment for compressional waves, seen from above. Diagram. Wave advances from left to 

 right to an observer in front. Lever displacements positive rearward. Ball displacements necessarily re- 

 versed . • 



VIII. Waves of Compression and Refraction. 

 25. Longitudinal Vibration. — With the ap- 

 paratus arranged as in Fig. 4, let the levers 

 all be raised at the front ends, so as quite to 

 disengage them from the front cam axle. 

 This being, therefore, out of action, the rear 

 or horizontal harmonic of 3" double ampli- 

 tude forward and rearward thrust is alone in 

 play, as shown in plan by the parallel lines 

 normal to the axis in Fig. 11. Now, let the 

 ball ends of the levers (eylets) engage the 

 long shanks (6") of a series of horizontal, 

 right-angled bell cranks, the equidistant 



rate to more than about If" in the rarefied 

 phases. 



The great advantage of an arrangement 

 of this kind from the kinetic point of view 

 is the direct evidence furnished that each 

 ball in the first instance is actually in S. H. 

 M., and that the phase difference between 

 balls is proportional to their distance 



*The reader should remember that Fig. 11 is seen 

 from above and that direction rearward in the trans- 

 verse harmonic (down in figure) is positive wave ve- 

 locity left to right in the machine, becomes right to 

 left in figure. Balls in front reverse their positive 

 motion, 



