20 Prof. H. W. Dove on a New Photometer, 



side of the free surface — and in a still more surprising manner 

 in the dazzling light which directly emerges from the perfectly 

 dark cylindrical jet of water which is strongly illuminated from 

 within, at the moment at which it separates in drops or is 

 interrupted by a shock, while a light held in the dark space 

 immediately behind passes through unchanged. Since with 

 the frequency of interruption the probability for the limiting 

 angle of total reflexion increases, we see directly how that which 

 is seen in a bubble in water, multiplying itself, produces 

 white foam — just as a crack in ice forms the transition to white 

 snow, a crack in glass explains the white powder of a broken 

 glass drop. But this powder is only white if the drop was of 

 colourless glass ; it is greenish if it were of green glass, just as 

 the froth of champagne is clearer than that of beer, of coffee, or 

 of chocolate, all, however, exceeded in whiteness by soap-bubbles. 

 The depolarization of polarized light incident on a snow surface 

 proves that the totally reflecting surfaces are inclined in all 

 directions ; and hence we can form an idea of the process of the 

 irregular reflexion of light from a white, grey, or coloured sur- 

 face — in the latter cases if, to the conditions producing the 

 first phenomena, we add that of absorption in the passage 

 through coloured substances. 



On this view the rough surface of an unpolished body will 

 not reflect light irregularly because it represents mirrors inclined 

 in all directions ; for on this supposition the colour of the body 

 would not be perceptible; on the contrary, it would present 

 surfaces to the light which facilitate the penetration at an almost 

 vertical incidence. Hence it is that increasing polish gradually 

 conceals the colour of a body. The fact that a very dark colour- 

 ing substance appears brighter in a state of powder than one in 

 a lump which had already a rough surface, does not disaccord 

 with this ; for here, by the greatly increased inequalities, the 

 reflexion at a very acute inclination is increased, whereas the 

 multifold rectangular incidences are little inclined to deepen the 

 colour in consequence of the great opacity. 



It follows from the preceding considerations, that diffused 

 light can be considered as if each point was self-luminous; in 

 this case if o is the aperture of the objective tube, and x the 

 angle which the diffusing surface makes with the axis of the 

 microscope, the surface which sends light upon the photographic 



picture will be — — , and the brightness will be proportional to 



sin cc 



this. If for two different substances with the same general 

 illumination compensation takes place at the angles x t and 



x n , their respective brightness will be in the ratio — and 



sin x I 



