1878.] the Photometry of the Magneto-Electric Light. 159 



with other persons the same uncertainty has been observed. Rumford's 

 method has never been a favourite, but I believe that if the following 

 modification of it be adopted, there can never be much difficulty in 

 comparing lights of similar colours. The problem which I set myself 

 was to obtain shadows of equal size, and to place them side by side as 

 seen by the eye. The following plan accomplishes this : — 



In an opaque screen of some half metre square two equal slits of 

 about half a centimetre in width and 7' 5 centimetres in length were 

 cut parallel to one another, the lengths being vertical. These slits 

 were separated by 7" 2 centimetres, and were covered with a diaphanous 

 screen, white tissue paper being apparently the best substance to 

 employ, though ground glass was also tried. The tissue paper was 

 placed at a small distance from the opaque screen, so that the texture 

 was invisible when the borders of the slit were sharp when viewed by 

 an observing telescope. A rod of a centimetre in diameter was placed 

 in front of the tissue screen, and so arranged that the shadows thrown 

 by the two sources of light entirely covered the slits. The movable 

 light was caused to travel along a scale fixed at such an angle as to 

 ensure this always occurring. The necessary precautions were taken 

 to ensure that the shadows falling on the screen were at equal dis- 

 tances from the centre of the rod. 



In order to cause the two shadows thus thrown to approach each 

 other, a divided lens of about 16 centimetres focus was made to form 

 the objective of a telescope, and by carefully separating the optical 

 centres of the two halves (see figure) when the line of division was 



kept in a horizontal position, the images of the two slits could be 

 caused to approach one another, and could be viewed in the same field 

 by an eyepiece. By turning the telescope in its Y's the same result could 

 be obtained, though in this case the image seen of the one slit weuld 

 be formed by the opposite half of the lens to that which formed it in 

 the first observation. Measurements were taken with standard lights, 

 till the adjustment was found which gave equal illumination of both 

 sides of the lens. When the opaque screen, the telescope, and the head 

 of the observer were carefully shielded from all extraneous light, the 



