TO THE MICROSCOPE. 133 



//^, and if the light happens to emerge from the section after 

 undergoing an exact number of cycles of changes, the object as 

 well as the neld will appear bright. If, however, the light emerges 

 in the middle of a cycle (Fig. 5 <?), where it is polarised along UU' 

 it will be quenched by the analyser, and the object will appear 

 dark on the bright ground. For intermediate thicknesses the light 

 is more or less reduced in intensity by the object, which therefore 

 appears somewhat darkened, or at any rate less bright than the 

 background. Now a slight diminution of brightness in a bright 

 field is not so conspicuous as even a faint illumination where the 

 surrounding field is perfectly dark, and for this reason, as every 

 microscopist knows, objects which are sufficiently polariscopic 

 to show up fairly well on a dark ground, frequently exhibit no 

 very noticeable polariscopic appearances when the field is bright. 



Comparison of the two phenomena.— It may perhaps be worth 

 while to examine a little more closely the analogy between the 

 oscillations in the pendulum experiment and the ether vibrations in 

 the beam of polarised light, and to enquire why the same cycle of 

 changes takes place in both. We have seen that the pendulum is 

 capable of permanently oscillating about the knot C in the plane 

 EE' of the strings, and that it is also capable of permanently 

 vibrating at a slightly slower rate about the points ^, ^ in a direc- 

 tion 00' perpendicular to that plane. When we pull the string 

 aside in a diagonal direction we set both of these motions 

 going at the same time, and the actual motion is a combination of 

 the two. [For in Fig. 5 ^ we may suppose the weight pulled aside 

 in the plane of the strings from P to E^ and then in a perpendic- 

 ular direction from E to /, and let go from /.■ the first of these 

 two displacements will start it oscillating in the direction EE'^ 

 and the second will start it oscillating in the perpendicular direction 

 00'\ But the quicker of the two oscillations gradually gai?is on the 

 slower. This gai?t is the cause of the cycle of changes in the ob- 

 served motion formed by the combinatio7i of the two oscillations. 



Whenever the pendulum has simultaneously performed an 

 exact number of complete oscillations in the direction EE^ and 

 an exact number along 00'., it will return to rest at /, and will 

 again begin to oscillate in exactly the same way that it did at the 



