OF ARTS AND SCIENCES. 29 



dicates that it is composed of light of a wave-length to which the 

 photographic plate is not very sensitive. 



An important source of error arises in almost all these measures 

 from the diffuse reflection from dust or scratches on the object-glass. 

 The eifect of this would be similar to an increased haziness of the 

 sky, and would tend to increase the apparent light received both 

 from Venus and the sky. In our measures this effect was reduced 

 to a minimum, as the object-glass had been cleaned shortly before the 

 transit, and the diffuse reflection was therefore very slight. A remedy 

 for this difficulty would be found by removing the object-glass and 

 substituting for it a minute hole. When the sky is hazy we should 

 expect an increased relative brightness near the edge of the sun. 

 This may account for the larger readings obtained by Mr. "Wendell, 

 as the sky was somewhat clearer during my observations than during 

 his. As the portion of the sky observed was only about 1' distant 

 from Venus, irregular clouds could not produce the observed difference 

 in light. In fact, the persistence of the phenomena under varying 

 conditions seems to leave little doubt that the disk of Venus was 

 really much darker than the sky near it. 



The solution of this problem would be greatly aided by researches 

 of the kind described below, a portion of which will probably be under- 

 taken at this Observatory. Measurement of the relative light of different 

 portions of the sun's disk and of the sky at various distances from it. 

 Measures of the sky at various distances from the moon, thus eliminat- 

 ing any effect corresponding to that of a solar atmosphere. Measures 

 of the light of the disk of the moon during the progress of a partial 

 eclipse of the sun. 



Spectroscopic Observations. 



The spectrum of the light received from Venus was observed with 

 a star spectroscope constructed by Hilger. The dispersion was such, 

 that about one third of the spectrum was visible in the field of view 

 at a time. Two prisms of dense flint glass were employed. This 

 spectroscope was attached to the large Equatorial, and was focused 

 on the limb of the sun. When Venus was brought upon the slit it 

 appeared as a broad band traversing the spectrum lengthwise, which 

 could be compared directly with the solar spectrum on each side of 

 it. The slit was then broujrht tangential to the limb of Venus, so as 

 to receive the light grazing its surface. The breadth of the dai-k band 

 was thus reduced and flickered owing to the slight unsteadiness of the 



