■1903-4.] Mr J. E. Milne on a New Form of Juxtapositor. 357 
appropriate openings for the entrance and exit of the light. The 
best place for the attachment of the apparatus to the spectro- 
photometer will, of course, vary to some extent with the pattern 
of the instrument — in the author’s case it is mounted immediately 
in front of the collimator slit. When the juxtapositor is so 
situated with regard to the spectrophotometer, the upper or “ com- 
parison ” beam of light enters face AB (fig. 2a) and meets the 
interface CD at an angle of 45°, and the part of it falling on the 
area OC is reflected upwards by the silvering. The other part, 
which falls on the unsilvered surface OD, passes straight on 
and out through the face CF, and is not used. In the same 
way the lower or “absorbed” beam enters face DE, and is 
reflected upwards by the silvering on the face EF, and the 
part of it incident on the lower half OD of the interface 
DC continues on its vertical course upwards. The other part, 
which falls on the silvered surface OC, is reflected out through 
the face CF, and is not used. The two beams which are re- 
spectively reflected and transmitted by OC and OD pass upwards 
in a common vertical direction, and have their edges in complete 
contact along a plane through OL normal to the paper. The 
beams thus brought into contact are reflected once more at the 
silvering on the face GH, and pass out through the face HC 
parallel to their original direction. 
In the above discussion the action of the juxtapositor has been 
explained in a particular case — namely, when attached to a spectro- 
photometer immediately in front of the collimator slit— and we 
