Gkeenhouse Notes. 



153 



strongly upon the plants, four of the front sides being supplied with 

 plain, orange, blue and red panes respectively, and the remaining side 

 being left open, with no screen whatever. The panes of glass used in 

 these screens were the best samples of their respective colors which 

 could be purchased. The blue, however, was really a purple, and the 

 orange was a dark amber. Spectrophotometric measurements were made 

 of these panes under the direction of Professor E. L. Nichols, by Miss 

 Mary C. Spencer, whose diagrammatic results are shown in the accom- 



A BC D E r C H 



rved leliov^ Green £>lue 



Spectro-Photometric Measurements of Panes. 



panying chart. The spectrum is represented below, by the familiar Fraun- 

 hofer lines and by the relative positions of the different colors. The 

 Fraunhofer lines are projected upwards by dots to show readily what 

 portions of the spectrum are cut oil bj the various samples. The 

 horizontal lines measure the amount of light of any color transmitted, 

 in decimal figures. Thus it will be seen that the blue glass transmitted 

 about ten per cent of the red rays, less than five per cent of the yellow, 

 while it was almost perfectly transparent to the blue rays. The red 

 pane transmitted nearly twenty per cent of the red rays, very few of 



20 



