20 EDWARD M. NELSON ON 



tion, will show that it is so, for the emergent beam is parallel 

 to the immergent one, and the box is in a line with the beam. 

 It is as easy to place this box between the lamp and the micro- 

 scope mirror as it is to place a common bull's-eye. 



In the figure the light is seen entering the circular hole in the 

 left-hand end of the box ; it falls on the mirror M, is reflected 

 on to the pile of plates G, and thence to the mirror M^, by 

 which it is reflected out of the hole in the right-hand end 

 of the box. 



The diameter of the hole is the diameter of the required 

 beam ; d, the breadth of the mirrors MM^, should be somewhat 

 {say 1/4 or 3/8 in.) greater than that of the hole. P is the 

 polarising angle ; a the inclination of the mirrors MM^. MM^ 

 the length of the mirrors, and G the length of the glass plates, 

 C is the base of the mirror triangle, shown in the figure. Then 



90° — P 



a = — - — ; M = dseca ; G = dsecP ; C = dtsma. 

 A 



The following are the inside measurements of the box : Length = 

 2 G — c\ width = d ; depth = 2 cZ -f- 1/2 in. For flint glass 

 /ot = 1-62 ; P = 58° 20' ; a = 15° 50' ; G= l-904cZ ; M = l-039c? ; 

 c = 0-284c?. Note, extra-thin slip glass is a suitable glass for 

 a pile of plates. The glasses should be put into strong sul- 

 phuric acid for a couple of hours, then washed in many changes 

 of water, then placed in methylated spirit, then dipped in 

 absolute alcohol and wiped clean. When treated in this 

 manner they will keep clean for a long time, otherwise they 

 will soon get steamy and sticky, and require frequent cleaning. 

 All microscope slides and cover- glasses should be cleansed in 

 this manner. 



No reference has yet been made to the two plano-convex 

 lenses in the lower compartment. The difference between an 

 instrument of this form and a Nicol will be seen when they 

 are rotated. In this instrument the emergent beam would 

 describe a circle of some inches diameter round the immergent 

 beam, while the displacement caused by the rotation of a Nicol 

 is slight, and can be adjusted by the substage centring screws. 

 Dr. Sylvanus Thompson, who has done so much for science by 

 his improvements of the Nicol, devised the following method 

 to meet this case. He said that if a 1/4 wave-length plate was 

 fixed with its axis at an angle of 45°, and if a similar plate 



