344 PROCEEDINGS OF THE AMERICAN ACADEMY. 



the oblique end of the box, and was thus reflected directly toward the 

 ground glass. The white-paper lining of the sides aided in securing 

 the desired diffusion of light. In this way a remarkably even illumi- 

 nation of the large sheet of ground glass was obtained. By means of 

 a square mat of opaque paper a clear field 41 cm. square was produced, 

 and this constituted the larger of the two areas of illumination used in 

 the experiments. The ground glass was first placed with the unground 

 surface toward the table, but later the ground side was turned in that 

 direction, in order to avoid the reflection of light from the small lamp 

 opposite. 



The lamp ( P^, sending down vertical rays, was essentially like the 

 small lamp (Sm), being similarly provided with a single 110-volt fila- 

 ment. There was arranged in front of it, however, a sliding diaphragm 

 (not shown in the figure), by means of which the size of the orifice, and 

 therefore the amount of light coming from it, could be regulated at 

 will. 



The filament of the small lamp (Sm) and the ground glass (g) of the 

 large lamp (L(j) were each exactly 2 meters from the central point (p) 

 of the table. For convenience this middle point may be spoken of as 

 the working point or wot'king posit/on. All light, except that coming 

 directly from the sources described, was excluded, as far as possible, 

 from the working position by means of diaphragms (d), screens of 

 heavy pasteboard or wood (s, s', s"), and black cloth (cs, cs', cs"). 

 The intensity of illumination of the ground-glass surface could now be 

 easily regulated by varying the distance of the lamp Lg' from the 

 oblique reflecting surface in the box Lg. With the aid of a Lummer- 

 Brodhun photometer placed at the working position (;;) it was an easy 

 matter, by varying the distance from o of the light Lg\ to make the 

 intensities of the two lights exactly the same. 



It is now necessar)'^ to compare the respective areas of the two sources 

 of illumination. The small filament was 1 mm. in diameter and 15 mm. 

 long, but at each end was a small knob where the platinum wires were 

 attached to the filament, and when the latter was at white heat these 

 glowed as well. Thus the total length of glowing surface was about 

 17 mm. Since the illuminated surface of the ground glass was 41 cm. 

 square, as has been stated, its area was 1()8,100 sq. mm. If in the 

 small lamp we regard the filament only, we shall have an area of 

 15 sq. mm., and therefore the ratio of the two areas will be that of 

 11,207 to 1. If, however, the length of the smaller light is considered 

 as 17 ram., the ratio becomes 9,888 to 1. Since the ratio cannot be 

 obtained with great accuracy, and the exact proportion is of only minor 

 importance, we may, for all practical purposes, regard the ratio as 

 10,000 to 1. 



