Diffusion Photometer, 27 



filled with diffused light, that part furthest removed dark 

 and obscured. The crack may be very minute, the effect is 

 the same. Only when the edge of the crack is so turned that 

 the source of light shines equally on either side of it, or if two 

 sources of light of similar colour illuminate separately and 

 with equal intensities the material at either side of the crack, 

 does the discontinuous appearance vanish. On this effect the 

 photometer to be described is based. 



If, in fact, two parallelepipeds of paraffin be cut, of equal 

 dimensions, and planed smooth so that they can be laid 

 accurately together on similar faces, it will be found that a 

 very sensitive apparatus is obtained, so that an appearance 

 of homogeneity is only secured by nice adjustment to a plane 

 of equal illumination. To compare two sources of light it is 

 sufficient to place the compound parallelepiped with its plane 

 of discontinuity at right angles to the line joining the sources 

 of light and cut by this line. Then, shifting the parallelepiped 

 between the lights, and regarding the fine line of division on 

 its surface, to find the point at which this line is no longer, 

 or only with difficulty, descernible. The distances are now 

 measured in the usual way, and the relative intensities of the 

 light reckoned as inversely as the squares of their distances 

 from the plane of juncture of the parallelepiped. 



In the case, however, of lights of dissimilar colour, the 

 appearance of the photometer is no longer uniform, but that 

 of two softly glowing substances having different shades of 

 colour. The difficulty now of judging when equilibrium is 

 obtained is one which must arise with all photometers which 

 are true to colour. So far, however, as my experience goes, 

 comparing the gas-flame with the candle, the comparison of 

 intensity is not possessed of any uncertainty, the line showing 

 with a minimum of distinctness when the brightness at each 

 side is the same, and the position of equilibrium being reformed 

 with great accuracy on repeated observation. In extreme cases 

 tinted glasses interposed in front of the flame might be used. 



The explanation of the sensitiveness of this photometer is 

 not far to seek. The surfaces being compared are in juxta- 

 position and simultaneously in the field of vision. If, now, 

 we concentrate all our attention close to the line of junction, 

 the least disturbance in the equality of the illumination will 

 be most favourably displayed to the perception. The light 

 entering the eye, too, is soft and perfectly uniform over the 

 surfaces regarded. Photometers in which a kind of visual 

 memory has to be exerted in the comparison of two images 

 removed some 8 or 10 centimetres from one another do not 

 possess the same sensibility. There is, also, a notable absence 



