338 ON LIGHT. 



portion, then, on its arrival would meet there, not the 

 other portion of the same wave to which it originally be- 

 longed, but one in advance or in arrear of that by either 

 a whole, a half, or any part of an undulation, or any 

 number of such, according to the extent of the difference 

 in the quality of the aerial contents of the column. Sup- 

 pose, for instance, the light from the two halves of the 

 column to differ in their time of arrival by i, 3, 5, or any 

 odd number of semi-undulations of the most luminous 

 or the yellow rays; these then would interfere and totally 

 extinguish each other, and the apparent light of the star 

 would undergo a great obscuration, assuming at the same 

 time a hue complementary to yellow ; i.e., dark purple : 

 and so for other rays. Now the constitution of the air 

 is so irregular such a perpetual mixture of masses of it, 

 differing in temperature and moisture, is continually go- 

 ing on under the influence of wind-currents, that such 

 differences as above supposed must be almost con- 

 stantly in progress, even within the narrow sj^ace of a 

 column no wider than the pupil of the eye, much more 

 in that corresponding to the aperture of a small telescope. 

 The scintillations, with their accompanying changes of 

 colour, are beautifully seen through an opera -glass 

 [jiot binocular)., especially if somewhat out of focus, in 

 which case the colours and the darkness are seen, as it 

 were, to run over the circular disc into which the image 

 is dilated in a very singular and capricious manner. If 

 a small circular motion be given to the glass, so as to 

 make the image of the star (when in focus) describe a 

 circle, this will be seen as a luminous circle (as when a 



