187 



fleeted light are generally very different from 

 the tints of transmitted light ; particularly 

 when the transmission takes place through a 

 great portion of fluid. If there were no absorp- 

 tion of rays, the transmitted light would be of a 

 colour complementary to that of the reflected 

 light; and in general we judge ill of transmitted 

 light, by filling with water a shallow glass with 

 a narrow aperture. In a river the colour of the 

 reflected light comes to us always from the 

 interior strata of the fluid, and not from the 

 upper stratum *. 



Some celebrated naturalists, who have exa- 

 mined the purest waters of the glaciers, and 

 those which flow from mountains covered with 

 perpetual snows, where the earth is destitute of 

 the relics of vegetation, have thought, that the 

 proper colour of water might be blue, or green. 

 Nothing, in fact, proves, that water is by nature 

 white, and that we must always admit the pre- 

 sence of a colouring principle, when water 

 viewed by reflection is coloured. In the rivers 

 that contain a colouring principle, this princi- 

 ple is generally so little in quantity, that it 

 eludes all chemical research. The tints of the 

 Ocean seem often to depend neither on the 

 nature of the bottom, nor on the reflection of 



* Newton, Opt. Lib. \, p 2, prop. 10, prob. 5. Delaval, 

 on the permanent Colours of opake Bodies, in the Manchester 

 Mem., 1789, vol. ii, p. 240. 



