47G 



PROCEEDINGS OF THE AMERICAN ACADEMY 



tram showed a maximum photographic intensity in the neighborhood 

 of the A line, whence it gradually faded away to line .94, where it 

 reversed, and became direct again near 1.12, where it disappeared. 

 This reversing action was noticed more markedly in the case ol' the 

 red glass, to be presently described. On the more refrangible side 

 of the A line there was a faint absorption band extending half-way 

 to D, and after that a uniform spectrum till D was readied. Here 

 it began to fall off, and soon disappeared. Between F and G there 

 was a small amount of lijilit transmitted. 



With a single red glass there were three maxima, the largest 

 between F and G, the next in size between A and D, and a small 

 one in the neighborhood of the line 1.12 micron. Between the last 

 two maxima there was a reversed band culminating in the neighbor- 

 hood of line .94. By the insertion of another red glass the maximum 

 between F and G was reduced to a small band in the vicinity of F, 

 and the reversed area was transferred somewhat lower down in the 

 spectrum, so that its maximum occurred near wave-length 1.04. 

 With the red and blue glasses, and with the asphalt, there was 

 apparently no reversed area. The former had three maxima, — at 

 F, just below D, and just above A. The last was the strongest 

 marked, and the one near D was very small. The asphalt had only 

 one maximum, and that was just below D. 



If the length of the exposure with two pieces of red glass be in- 

 creased, the limits of the reversed area will advance in both directions, 

 as is shown by the figure, where abscissas represent the wave-lengths 

 in the prismatic spectrum, and ordinates are proportional to the loga- 

 rithms of the exposures. The shaded area shows the darkened 



1.82 



1.38 1.12 .91 



.70 

 A 



.59 

 D 



