304 



Sir W. de W. Abney. On the Variation in 



tion from Mi remained the same for all rays. The star-shaped opaci- 

 ties were then introduced into the beam X at D, and when necessary, 

 B was rotated with known and fixed apertures, and the patches of 

 light again made equally bright by means of A. It was found that 

 the apertures of A varied as the different spectrum colours passed 

 through the deposits, forming the graduated star. Using the same 

 scale for the spectrum as used in my former papers (B is 61*3. 

 Li 59-7, C 58-1, D 56, E 39-8, F 30'05, Li 22-8, G 11*2), the absorp- 

 tions were calculated for the whole spectrum. It was found that the 

 coefficient of absorption (obstruction) of white light and of the ray 

 26*8, coincided, and taking this as unity (for a purpose which will be 

 seen presently) the coefficients of the other rays are as follows : — 



Table I. 



Scale number. 



Absorption. 



59 to 49-8 



0-87 



47 '5 



0-90 



42 -9 



0-92 



38 3 



0-93 



337 



0-95 



29-1' 



0-97 



26-8 



l'OO 



22 -2 



1-02 



17 6 



1-02 



8 -4 



1-08 



The transparencies of the different parts of the star to lamplight 

 were measured and calculated out in powers of - 2, the light trans- 

 mitted through the part on which no deposit appeared being taken as 

 zero. The following are the transparencies as calculated : — 



Table II. 



Opacity. 



Transparency in 

 powers of — 2. 



l\ T o. 1 







2 



0-38 



" 3 



0-75 



>, 4 



105 



„ 5 



1 -73 



„ G 



2 -36 



„ 7 



3 5 



,. 8 



4-16 



., 9 



52 



„ 10 



5-9 



„ 11 



6-9 



„ 12 



8 -9 



