100 COSMOS. 



of commotion), destroy each other by inequality of path; 

 that the light of one ray added to the light of the other 

 produces darkness. When the retardation of one system of 

 waves in reference to the other amounts to an odd number 

 of semi-undulations, both systems endeavour to impart simul- 

 taneously to the same molecule of ether equal but opposite 

 velocities ; so that the effect of their combination is to produce 

 rest in the molecule, and therefore darkness. In some cases, 



On pushing in the eye-piece, the disc of the star increases 

 in diameter and a dark point appears in its centre ; when the 

 eye-piece is made to recede still further into the instrument, 

 a luminous point will take the place of the dark point. On 

 causing the eye-piece to recede still further, a black centre 

 will be observed. If while the centre of the image is black 

 we point the instrument to a star which does not scintillate, 

 it will remain black as before. If, on the other hand, we 

 point it to a scintillating star, we shall see the centre of the 

 image alternately luminous and dark. In the position in 

 which the centre of the image is occupied by a luminous point, 

 we shall see this point alternately vanish and reappear. This 

 disappearance and reappearance of the central point is a 

 direct proof of the variable interference of the rays. In order 

 to comprehend the absence of light from the centre of these 

 dilated images, we must remember that rays regularly refracted 

 by the object-glass do not reunite and cannot consequently 

 interfere except in the focus ; thus the images produced by 

 these rays will always be uniform and without a central point. 

 If in a certain position of the eye-piece, a point is observed 

 in the centre of the image, it is owing to the interference of 

 the regularly refracted rays with the rays diffracted on the 

 margins of the circular diaphragm. The phenomenon is not 

 constant, for the rays which interfere at one moment no 

 longer do so in the next, after they have passed through atmos- 

 pheric strata possessing a varying power of refraction. . We 

 here meet with a manifest proof of the important part 

 played in the phenomenon of scintillation by the unequal 

 refrangibility of the atmospheric strata traversed by rays 

 united in a very narrow pencil." 



