on the Optical Phenomena of the Atmosphere. 13 



tion in which the observer recedes from the diaphanoraeter must 

 be perpendicular to the surface of the latter, as otherwise the 

 quantity of light reflected would be very different in different 

 experiments *. 



Table of observations with Saussure's Diaphanometer, Circle a 

 1 inch, Circle A 1 foot in diameter. 



The column for Circle a contains the distances at which the 

 small circle disappeared ; the column Circle A, the distances at 

 which the large circle vanished. In the two following columns 

 stand the calculated angles under which the respective circles 

 disappeared. The last column (Q) contains the ratio of the re- 

 spective distances, — a number which, as already remarked, for a 

 perfectly transparent medium ought to be =12, but in the pre- 

 sent case is always less. In Nos. 4 and 6, though only one 

 distance could be measured, the angle under which the objects 

 disappeared was of interest. 



In general the quotient increases with the height, i. e. the 



* Compare Biot, Traite de Physique, vol. iv. p. 776. Black marble 

 reflects, according to Bouguer, 600 rays of every 1000 under an angle of 

 3° 35' ; under an angle of 30°, 50 of every 1000. Under an angle of 0°-3, 

 white marble reflects 721; under an angle of 2 0, 3, 614; and under an 

 angle of 15°, 211 rays. 



t Notwithstanding this great transparency, no stars were visible. 



X This number is in reference to the rocks 10 feet broad, mentioned at 

 No. 4. 



