564 CAPTAIN ABNET AND MAJOR-GENERAL TESTING 



With green 62, blue 110, red 180, he matched 92 white. 



His areas for the above would have been 60, 19, 112 respectively, making a total of 

 191. The area for the white would be 194. 



Again, with green 138, blue 94, red 128, he required white 10G. The areas would 

 have been as follows : 133'.}, 16, and 77 respectively, or a total of luminosity 226'5. 

 The luminosity of the white used was 223. 



Again, with blue and green alone, he required 288 green and 72 blue to match 141 

 white. The areas of the above were 281 and 12 '4 respectively, or a total luminosity 

 to him of 29 3 '4. The luminosity of the white used was 29 7 '4. 



XL. Comparison of Gas Light, Sky Light, and the Electric Light. 



Fig. 24 (Plate 2) shows the proportions of the different rays in sky light, gas light, 

 and the electric (crater) light, the last being taken as the standard of comparison. 

 The sky measured, it may be stated, was a fairly blue sky and not very pale. The 

 light diffused through a cloud on a cloudy day in April we have found to be almost 

 exactly similar to the light of the electric arc, and, in fact, is degraded sun light. 



We have shown how the three coloured sectors vary in proportion to form a grey 

 when examined by electric light and by gas light. The variation would be even greater 

 when sky light of the blue shown was employed. On a cloudy day, however, the 

 proportions of each colour would be approximately the same as found for the electric 

 light. It is evident that for quantitative measures for testing light no reliance 

 can be placed on the results, unless light of an uniform character be always employed. 

 Day light, being composed of variable amounts of sun light and sky light, should, in all 

 cases, be avoided. 



XLI. Reflection of Light from Metals. 



As a matter of curiosity, we wished also to determine the intensities of the different 

 rays reflected from some of the metals which were coloured. We took ordinary 

 polished copper, such as is supplied in commerce for etching purposes ; a piece of the 

 same copper highly burnished ; a piece of highly burnished brass ; and a piece of highly 

 burnished gold. Fig. 27 (Plate 2 1) and the following tables give the result. In the 

 figure the continuous lines show the luminosity curves, and the dotted lines the 

 intensity curves. 



