468 THE POPULAR SCIENCE MONTHLY. 



rosy or purplish-pink, and at last the pure blue of the sky. The colors 

 were wonderfully vivid for the time of day, although, of course, not 

 60 brilliant as those of a well-developed sunset ; but it unfortunately 

 seems to have very generally passed unnoticed. Inquiry among my 

 neighbors failed to discover any one who had seen it. 



Numerous observations in many parts of Europe and this country 

 leave little room for question that the corona is produced in the upper 

 atmosphere, and that it was continuously present above the cloudy or 

 dusty lower air over a large part if not over the whole of the earth 

 throughout 1884 and 1885. 



The explanation of the optical process by which such a corona may 

 be produced offers no particular difficulty. It is a relatively simple 

 effect of diffraction, an effect of the same nature as that seen in the 

 colored rings surrounding a light looked at through a glass that is 

 faintly frosted over, as may be noticed almost any cold -winter evening 

 when looking out of a window. A brief statement of the process may 

 be made, following the explanation given by Kiessling, to whom the 

 author is much indebted in the preparation of this article. 



Let us first consider the action of a beam of parallel rays of mono- 

 chromatic light — that is, of strictly one-colored light, whose waves all 

 agree in their period of vibration — as it passes an excessively fine 

 thread stretched at right angles to its path, and falls on a screen be- 

 yond. The waves will be turned aside from and bent around both 

 sides of the thread, as if diverging there from new centers of radia- 

 tion. This is diffraction. A gross figure of the process is here given 

 (Fig. 1) on a plane at right angles to the thread, T H. The point A 



on the screen will be illuminated, although it is behind the thread, for 

 the waves that reach A from either side of the thread agree in phase. 

 Take a point, B or C, such that the distance B H exceeds B T by half 

 a wave-length. Then the diffracted waves which agreed in phase at 



