﻿300 M. H. Wild on the Absorption of Light by the Air, 



but also as illuminators, while when air or liquids are enclosed 

 in tubes their absorptive action alone comes into play. 



The unexpected magnitude of the absorption of the air as de- 

 duced from my measurements led me to investigate whether the 

 transparency of the air for different colours could not be experi- 

 mentally determined in an analogous manner. With this view the 

 relative brightness of the two screens while at a distance of 30 me- 

 tres was determined — in one case for red light by placing a deep- 

 red glass in front of the ocular of the photometer, in the other 

 case for blue light by the use of a combination of a cobalt with a 

 bluish-green glass. Two sets of observations made in this man- 

 ner, on the 9th and 14th of September, showed that red light is 

 absorbed less strongly than blue, or that the coefficient of transpa- 

 rency of the air for red light is greater than for blue. I do not 

 give the numerical values for the coefficients, as they are only very 

 approximately determined, and may be essentially modified by 

 later, more accurate, and more complete observations. 



I was prevented by pressing business from resuming these 

 observations; and only in August 1867 could I proceed to attempt 

 to make them by a new method of observation, which was more 

 in accordance with that which I had first chosen. With the deli- 

 cacy of my photometer and the unexpected magnitude of the 

 absorption of the air, it seemed that I ought to be able to de- 

 monstrate it by means of tubes of about 2 metres in length. 



The apparatus which I used consisted of an oiled transparent 

 paper disk of 30 centims. diameter, which was placed near a 

 window only illuminated by diffused daylight, and which, in 

 order to obtain as uniform illumination as possible, was rotated 

 by an axis passing through its centre which was moved by clock- 

 work. At right angles to the face of this disk, on a horizontal 

 diameter of it, and with their centres opposite, two horizontal 

 sheet-metal tubes were placed of 2*4 metres length and 0*1 

 metre diameter. They were provided throughout their entire 

 length with twelve diaphragms, each with an aperture of 0*06 

 metre, and were moreover closed at the ends with plate glass. 

 Each had two smaller lateral apertures, which in the case of 

 one tube were constantly open ; while one of the apertures in 

 the second was closed, and the other connected with an air-pump 

 by means of a caoutchouc tube with an enclosed spiral. The 

 photometer was placed in front of the ends of the tubes furthest 

 from the paper disk, in such a manner that the light from the 

 two halves of the paper disk could pass through both tubes into 

 the two apertures of the prism apparatus. The chief difficulty 

 and uncertainty in this form of the experiment arises from its 

 being almost impossible to keep the two halves of the disk at a 

 constant ratio of brightness. I could only effect this for the 



