Comparison of Sources of Artificial Illumination. 105 



We shall have further occasion presently to compare these 

 two methods, and to place the results side by side with those 

 obtained by the use of still other forms of photometer. 



The Light of the Electric Arc. 



Our experiments with the arc-light were made for the most 

 part upon a lamp of the " long-arc " type, with about 10 

 amperes of current and 50 volts. The lamp was suspended at 

 a distance of 70 cm from the block of magnesium^carbonate. It 

 could be raised or lowered at will, so as to expose the latter to 

 the rays emitted in the horizontal plane, or at angles above or 

 below the same. A lens placed near the lamp threw an 

 enlarged image of the carbons upon a screen carrying a vertical 

 scale, so that the length of the arc could be determined at a 

 glance. 



Measurements were made with the magnesium-carbonate in 

 the horizontal plane with the arc, and at positions 16° above, 

 and 14°, 28° and 50°, respectively, below that plane. The 

 carbons used in the lamp were l*2 cm in diameter. They were 

 of the commercial brand known as the " Faraday carbon," and 

 were copper plated. 



The observations made upon rays given off in the horizontal 

 plane and in directions below the horizon, viz : in the three 

 last named positions above mentioned, showed that although 

 the quantity of light increased rapidly from the horizontal 

 plane downward, reaching a maximum in the neighborhood of 

 50°, as in all arc-lamps of the type in question, the quality of 

 the light was very nearly constant throughout that entire zone. 

 The variations in color with change of angle were indeed very 

 much less than those due to the fluctuations occurring during a 

 single set of readings in a single position. Just above the 

 horizontal plane, however, there occurred a quite abrupt and 

 very marked change in the color of the rays emanating from 

 the arc, and the light became much bluer than that within the 

 zone below the horizon. 



This change of color, which was apparent to the unaided 

 eye, was found to be due chiefly to the increased brightness of 

 the wave-lengths lying beyond the E line ; the distribution of 

 intensities in the red, and yellow being nearly the same in the 

 spectrum of the rays falling below the horizontal plane and in 

 that of the light 16° above the horizon. The manner in which 

 the relative brightness of the spectrum increases toward the 

 violet is best seen by an inspection of curves III and IV, fig. 2, 

 in which these results are exhibited. 



