ON SCHOOL-BOOKS AND EYESIGHT. 237 



specimen is viewed from either end, and indicates s-=o, the total 

 brightness consisting only of diffused brightness d. 



If a sheet of paper is tested in this way, and is observed through 

 eye-tube a, the brightness of the half illuminated by reflection from the 

 mirror B will be due to the sum of the diffused brightness d and of the 

 partial specular brightness y. The brightness of the half illuminated 

 by reflection from the mirror A is due to diffused brightness only. The 



a\ 



/ 

 / 

 / 

 / 

 / 

 / 

 / 

 / 

 / 

 / 

 / 

 / 



\ , '^ Specimen under Test 



Ooub/e 

 Prism 



^ -y^ ^ 



, , I Z 3 4- 



t- —I Ooubte Prism in P/an srA/f t= i i i — I incn£S 



Gloss Testing. 



lamp must therefore be brought nearer the nairror A to make the 

 brightness of the two halves equal. If, when a balance has been 

 obtained, the pointer stands at 66' 6, this means that the bright- 

 ness viewed from a direction nearly parallel to the direction of observa- 

 tion consists of brightness s due to specular reflection, which is twice 

 the brightness d due to diffused reflection, and, therefore, 66-6 per 

 cent, of the total brightness. As a check, an observation may be made 

 through the other eye-tube. 



When the specular brightness is more than 70 per cent, of the total 

 brightness an imperfect image of the source of light begins to be formed. 

 A bright glistening patch appears on one half of the specimen, and it 

 becomes difficult to match this accurately with the brightness of pure 

 diffused reflection. For this reason the method in its present form is 

 not applicable to materials having a pronounced gloss. 



In adjusting the instrument it is laid on a mirror. The lamp is 

 seen by reflection on one side of the partition, and when the lam.p 

 is moved it ought to appear to approach to or to recede from the eye in 

 a straight line. The instrument should be used in a darkened room. 

 The constancy of the candle-power of the lamp is immaterial ; but it is 

 advisable to use an evenly frosted bulb, as otherwise the asymmetry of 

 the filament may introduce errors. The surface undergoing test shoultj 

 be quite flat. 



