IN ITS RELATION TO THE EYE. 461 



the effects on the eye, as determined in preliminary experiments, 

 differed so little from those gotten from the blown white glass as to be 

 considered as of little significance for the present work. These re- 

 flectors were all used with 2^4 in. form " H " holders, and were in- 

 stalled on the ceiling pendant in accord with the principles of direct 

 lighting. Full- frosted tungsten lamps were used as light sources 

 with each installation. The results of this series of experiments are 

 represented in Chart V.- 



In the tables^ referred to in footnote i, p. 445, we have again 



2 In considering these results it should be borne in mind that these re- 

 flectors have been used to produce certain variations in illumination effects 

 and that the work has not been conducted as a specific test of reflectors. For 

 example, in order to secure in all cases approximately equal illumination at 

 the test object, the lamps had to be operated at slightly higher voltages for 

 some reflectors than for others. This produced for the different reflectors 

 slightly different relative brightness values for outer surface and opening 

 than would have been obtained had the lamps all been operated at the same 

 voltage. Also clear and bowl-frosted lamps are more commonly used with 

 these reflectors than full-frosted lamps. One effect of using clear or bowl- 

 frosted lamps with them in this work would have been to have increased the 

 brightness of both opening and outer surface of the reflectors and to have 

 given, there is good reason to believe, a correspondingly uniformly poorer 

 result for the eye. It is never entirely safe to predict results under conditions 

 differing even slightly from what have been used, but from data at hand there 

 is no reason to think that the change would have produced any significant 

 difference in the relative rating of these reflectors. The full-frosted lamps 

 were used for two reasons: (a) to test the whole group of reflectors under 

 conditions as favorable as possible for the eye. This is admittedly only one 

 point of view ; the results might have had a more direct practical bearing had 

 clear or bowl-frosted lamps been used. And (5), which is the chief reason, 

 for the sake of making the work as far as possible comparable with the pre- 

 vious work, we desired to make the illumination of the test object as nearly 

 equal as could be for the different reflectors, translucent and opaque, and 

 equal to that used in the former work. This was best accomplished by the 

 selection of lamps made. 



3 The following points might perhaps be cited in connection with the 

 brightness specifications given in these tables. In case of the translucent re- 

 flectors, installed pendant, two important items of surface brightness should 

 be taken into account, the brightness of the opening and the brightness of the 

 outer surface of the reflector. If a dense reflector is chosen, for example, 

 the brightness of the opening tends to become excessively high; also its ap- 

 parent or physiologic brightness is increased by induction from its dark sur- 

 roundings which effect does not register on the photometer. If, on the other 

 hand, the reflector chosen transmits too much light the brightness of the outer 

 surface of the reflector becomes too high for the comfort and welfare of the 



