October 15, 1915] 



SCIENCE 



535 



Wava Length in Microna (m) 



Diagram showing the Distribdtion of Energy 

 IN THE Visible Spectrum op Sunlight and of 

 THE Light prom a Nitrogen-filled Tungsten 

 Lamp; also of tlie lamp-light filtered through co- 

 balt glass and through the daylight glass 

 (G.172 CD). 



The accompanying diagram shows the dis- 

 tribution of energy in the spectrum of sunlight 

 and of the nitrogen-filled tungsten lamp; it 

 also gives the curve of the light from the 

 nitrogen-filled tungsten lamp filtered through 

 cobalt glass and through the daylight glass, 

 1Y2 CD. 



The curve for the light filtered through the 

 daylight glass approximates very closely to 

 tliat for sunlight, especially between wave- 

 lengths .45ju, and .65;n, that is, in the region of 

 the visible spectrum giving the greatest amount 

 of useful light. 



While light filtered through the proper thick- 

 ness of cobalt glass may look white to the eye, 

 it gives even more imperfect color values than 

 the unfiltered artificial light. This is intelli- 

 gible from the violent irregularities of the 

 curve of the light through cobalt glass, and 

 especially the enormous excess in the red as 

 shown in the diagram. 



Light filtered through the daylight glass has 

 been very critically tested in my laboratory 

 with microscopic objects stained with many 

 different dyes, and some of them with several 

 dyes on the same specimen to differentiate the 

 various structural details in the same organ. 

 To make sure that the microscope modified in 

 no way the color values, apochromatie objec- 

 tives and compensation oculars were used as 

 well as the achromatic objectives and Huy- 

 genian oculars. 



The tests were made in the daytime by a 

 window so that it was possible to turn the 

 mirror from the artificial daylight to true day- 

 light instantly and to determine any difference 

 in appearance, if such difference existed. It 

 was impossible to detect any difference in the 

 colors, although tests were made with the most 

 varied specimens and with a full range of ob- 

 jectives, including the 1.5 mm. oil immersion. 



Not wholly trusting my own judgment, I 

 secured that of colleagues in histology and 

 embryology and microchemistry from our own 

 and five other institutions, and their judgment 

 entirely confirms my own. 



In practise it was found desirable to have 

 the daylight glass finished with the ground or 

 velvet surface on one or both sides, and to 

 place it in the opening of an opaque screen be- 

 tween the artificial light and the microscope. 

 With this arrangement of the light, the effect 

 is like that from a white cloud. 



As stated above, this glass filter was designed 

 to give daylight qualities to the light from a 

 nitrogen-filled tungsten lamp, and gives the 

 most perfect and satisfactory illumination for 

 the microscope with this lamp. To thoroughly 

 test the glass, the other light sources used in 

 microscopic work were also tried; viz., the 

 vacuum tungsten, and the carbon-filament 



