LUMINOUS SURFACES. 



lO/ 



circles of diffusion increase relatively the dimensions of small objects much 

 more than of large ones, with regard to which, the dimensions of the small 

 circles of diffusion which the eye furnishes, when properly accommodated, 

 become insensible. 



I. Luminous surfaces appear larger. We can never judge exactly of 

 the dimensions of a slit or small hole through which a bright light escapes ; 

 it always appears to us larger than it really is, even with the most exact 

 accommodation. Similarly, the fixed stars appear in the form of small 

 luminous surfaces, even when we make use of a glass which allows of perfect 

 accommodation. If a gridiron with narrow bars the spaces intervening 

 being exactly equal to the thickness of the bars is held over a light surface, 

 the spaces will always appear wider than the bars. With an inexact accom- 

 modation, these phenomena are still more remarkable. Fig. 103 exhibits 

 a white square on a black foundation, and a black square on a white founda- 



Fig. 103. Experiment i. 



tion. Although the two squares have exactly the same dimensions, the 

 white appears larger than the black, unless with an intense light and an 

 inexact accommodation. 



2. Two adjacent luminous surfaces mingle togetlter. If we hold a 

 fine metallic wire between the eye and the sun, or the light of a powerful 

 lamp, we shall cease to see it ; the lighted surfaces on all sides of the 

 wire in the visual range pass one into the other, and 



become mingled. In objects composed of black and 

 white squares, like those of a draught-board (fig. 1 04), 

 the angles of the white squares join by irradiation, 

 and separate the black squares. 



3. Straight lines appear interrupted. If a ruler 

 is held between the eye and the light of a bright 

 lamp or the sun, we perceive a very distinct hollow 

 on the edge of the ruler in the part corresponding 

 to the light. When one point of the retina is 

 affected by a light which undergoes periodical and 



regular variations, the duration of the period being sufficiently short, there 

 results a continuous impression, like that which would be produced if the 

 light given during each period were distributed in an equal manner through- 



Fig. 104. Experiment 2. 



