OF ARTS AND SCIENCES. 



27 



in this way constitute a set, and give the relative light of the sky and 

 Venus. To eliminate any difference in the prisms the photometer 

 was rotated 180° after each set, but no perceptible difference is indi- 

 cated in this way. To reduce the observations, the first reading was 

 subtracted from the second, and the third from the fourth. Calling 

 the sum of these differences A, the relative light, L = tan^ i A. It 

 will also be convenient to use a method of expressing the light in 

 stellar magnitudes, according to the method already used in this Ob- 

 servatory for comparing nebulae and portions of the moon. When 

 surfaces are thus compared, portions of equal area are selected and 

 reduced to stellar magnitudes by the formula of Pogson. We shall 

 then have the difference in magnitude, 31^ 2.5 log L. In Table III, 

 the successive columns give a current number, the Cambridge mean 

 time, the difference in light of equal areas of Venus and the sun 

 expressed in stellar magnitudes, the corresponding quantities for the 

 sky near the edge of the sun, and these same ratios expressed in per- 

 centages, that is, assuming the light of the centre of the sun equal to 

 one hundred. The last column gives the initial of the observer. 



TABLE III. — Photometric Observations. 



The result for the light of the sky in the first line of the table 

 depends upon eight settings. 



These observations show a well-defined increase in light of the sky 

 near the edge of the sun as compared with that received from Venus. 

 This effect also seemed to me to be very perceptible without the pho- 

 tometer. To confirm it, I asked Mr. Wendell which looked to him 

 the brighter. He satisfied himself that Venus certainly appeared 

 darker than the sky. A slight difference was to be expected, since 

 there are instances on record of the visibility of Venus before first 



