86 Mr. A. P. Trotter on a New Photometer. 



screen has strips 5 millim. wide, and a third has 48 strips 

 3 millim. wide. White foolscap paper was used, soaked in 

 hot water to remove the glaze ; it was mounted whole, and 

 cut on glass when dry, the knife being held slanting as in 

 mount-cutting, to bevel the edge. These screens I find, and 

 my opinion is confirmed by friends who are for the most part 

 unfamiliar with photometers, to be easy to use. 



The advantage of vertical strips is that perfect symmetry 

 is attained, a condition which I find to be important in pho- 

 tometry. When perforations are used, dark holes on a 

 light ground are seen at one end of the screen, and light 

 holes on a dark ground are seen at the other end. The 

 balance is indicated by a strip at which these two patterns 

 become confused. I have not succeeded in getting a definite 

 easily-bisected strip of uniform tint with^ such screens. 

 When, however, strips are used, the spaces being equal to the 

 width of the strips, the appearance of the screen is perfectly 

 symmetrical, and the effect is improved by the use of a 

 diaphragm cutting off from the observer the edges of the 

 screens. When lights of similar colour are compared, the 

 light strips and dark spaces at one end are almost indis- 

 tinguishable from the dark strips and light spaces at the 

 other end. With 8 c. p. lamps 3 metres apart, and with the 

 8 millim. strips, one strip or one space can generally be made 

 to disappear ; that is to say, its edges become invisible when 

 examined from a distance of 6 or 8 feet. At a shorter 

 distance, one eye only must be used. With the 3 millim. 

 strips, a nearly uniform grid is seen, paling a little towards 

 the middle, and showing a band about 15 millim. wide, 

 including, say, three strips which are indistinguishable from 

 the two spaces between them. When lights differing in 

 colour are used, the narrow strips are distinctly preferable : 

 disappearance is nowhere complete, but the differently 

 coloured strips and spaces seem to blend at a definite place on 

 the screen. 



As screens 30 centim. long and requiring a box about 

 50 centim. by 30 might be considered cumbersome, I have 

 made a photometer with screens 22*5 x 3" 7 centim. effective ; 

 these are contained in a box measuring about 22 x 32 x 6 

 centim. deep. The diaphragm facing the observer is 15 x 

 3*5 centim. The screens are placed at 45° to the lights and 

 to the observer. I have tried other inclinations, but there is 

 no difficulty in getting a surface sufficiently dead-white to 

 work well at this angle. 



I find that a distance of 3 metres between two 8 c. p. glow- 

 lamps is rather great for easy reading of the photometer. A 



