498 Proceedings of Royal Society of Edinburgh. [sess. 
viewed from the front by using an opaque white screen in a 
darkened room ; or still another method is to remove the screen 
altogether, and replace it by the eyepiece of a telescope. The 
third plan becomes necessary when the light is faint, but the use 
of the eyepiece requires the head to be held strictly in one position, 
Which is fatiguing. From this drawback the use of a screen is 
free. 
After having for some time been engaged in the perfecting of an 
experimental model which I had constructed, my attention was 
drawn to the fact that an apparatus based on the principle of the 
divided lens had already been suggested by Schottlander.* His 
proposal, however, which differs from mine in its method of 
application, never seems to have been actually carried out. 
The main advantages of the above method are as follows : — 
1. The method is one for completely eliminating the dark space, 
not merely for minimising it, as in Hii filer’s method. 
2. The device is quite extraneous to the spectroscope, and the 
latter requires no alteration. The devices of both Brace, and 
Luinmer and Brodhun, require two spectroscope collimators ; 
while the device of the former requires in addition a special 
spectroscope prism. 
3. The two beams of light are subject to precisely the same 
conditions from the time of their entering the spectroscope slit 
till their ultimate emergence on the viewing screen. The only 
exception to this that can be urged is, that some difference might 
exist between the two half-lenses, which transmit the two beams 
respectively. It is, however, to be borne in mind that these are 
simply two halves cut from the same lens, so that any difference 
between them is very unlikely. Moreover, in all examinations 
where the light enters the collimator slit in two parallel beams, 
this possibility of error would be present if the collimator lens 
were not perfectly the same in its upper and lower portions, what- 
ever the form of spectroscope employed, and yet no investigator 
with whose work I am acquainted mentions such a source of error. 
4. The spectroscope prism can be adjusted to the minimum 
deviation for each colour. 
5. By the aid of this device the visual intensity of the 
* See Zeitschr. f. Instrk., ix. 98, 1889. 
