ON THE THEORY OF OPTICS. 411 



be always greatest when the difFereace of the optica) or refractive density of 

 the two substances is greatest. Thus the reflection from the common sur- 

 face of glass and water' is much weaker, than from a surface of glass ex- 

 posed to the air. Metals in general reflect a great proportion of the light 

 falling on them, and even the reflection from the common surface of glass 

 and mercury appears to be but little weaker than the reflection from the 

 surface of mercury immediately exposed to the air, so that the optical den- 

 sity of the metals must be exceedingly great. 



It appears also that a portion of the light falling on a reflecting surfiice 

 is always transmitted, at least to a certain depth, notwithstanding the appa- 

 rent opacity of any large masses of the substance. Thus, if we cover a small 

 hole of a window shutter with the thinnest leaf gold, we shall find that it 

 transmits a greenish light, which must have passed the reflecting surface, 

 but which, if the gold had been but one ten thousandth of an inch in thick- 

 ness, would have been wholly intercepted, and probably almost in the same 

 manner as by passing through 700 feet of water. In transparent substances, 

 however, the greater part of the light penetrates to ail distances with little 

 interruption, and all rays of the same kind, thus transmitted by the same 

 surface, form with the perpendicular an angle of refraction which is ultimately 

 in a certain constant proportion to the angle of incidence ; that is, for instance, 

 one half, three fourths, or two thirds, according to the nature of the sur-r 

 ffice. Thus, if the refractive properties of the substance were such, tliat an- 

 incident ray, making an angle of one degree with the perpendicular, would 

 be so refracted as to make an angle of only half a degree with the same line, 

 another ray, incident at an angle of two degrees, would be "re- 

 fracted, without sensible error, into an angle of one degree. IJut Avhen the 

 angles are larger, they vary from this ratio, their sines only preserving the 

 proportion with accuracy: for example, if the angle of incidence at the sup- 

 posed surface were increased to 90°, the angle of refraction would be 30" 

 only, instead of 45". Rays of the same kind are in general distinguished 

 by the same colour, although some rays, Avhich differ from each other in 

 refrangibility, have scarcely a discernible difference of colour; and it is 

 possible, on the other hand, to find a surface at which the ratio of the angles 

 is the same for rays of all kinds. (Plate XXVI. Fig. 369, 370.) 



