98 ON THE RELATION OF OPTICS TO PAIXTINIK 



black, lighted by the sun, is not less than balf as 

 bright, as shaded white in the brighter part of a 

 room. 



On the picture of the moon, the same white which 

 has been used for depicting the Bedouins' garments 

 must be used for representing the moon's disk, and its 

 reflection in the water ; although the real moon ha a 

 only one fifth of this brightness, and its reflection in 

 water still less. Hence white garments in moonlight, 

 or marble surfaces, even when the artist gives them a 

 grey shade, will always be ten to twenty times as bright 

 in his picture as they are in reality. 



On the other hand, the darkest black which the 

 artist could apply would be scarcely sufficient to repre- 

 sent the real illumination of a white object on which 

 the moon shone. For even the deadest black coatings 

 of lamp-black, black velvet, when powerfully lighted 

 appear grey, as we often enough know to our cost, when 

 we wish to shut off superfluous light. I investigated 

 a coating of lamp-black, and found its brightness to 

 be about ^fa that of white paper. The brightest 

 colours of a painter are only about one hundred times 

 as bright as his darkest shades. 



The statements I have made may perhaps appear 

 exaggerated. But they depend upon measurements, 

 and you can control them by well-known observations. 

 According to Wollaston, the light of the full moon is 



