

Hilfner Type of Spectrophotometer. 483 



from one part of the spectrum to another is effected by 

 rotation of the prism by a micrometer screw. 



The light source, A, is fixed to the cast-iron bed of the 

 instrument. The light passes partly above and partly through 

 a polarizing prism 0. The upper beam passes first through 



the substance whose absorption is 

 to be measured and then through 

 a thin wedge of neutral- tint glass, 

 by the lateral translation of which 

 the two beams may be made equal 

 in intensity. Both beams then pass 

 through the rhomb of glass B (fig. 5) by means of which the 

 two beams are brought into close juxtaposition. Immediately 

 behind this rhomb is the slit of the spectroscope, the edge of 

 the rhomb and the jaws of the slit both being in focus simul- 

 taneously when one observes them in the telescope. 



The light after collimation in the usual way then under- 

 goes dispersion by the constant-deviation prism, and passes 

 on to the second polarizing prism G, and so on to the 

 telescope. The second polarizing prism is mounted on a 

 circle H divided into degrees and reading by a vernier to 

 minutes, and by rotation of this the polarized beam of light 

 can be reduced in intensity in the usual way till the intensity 

 of the top and bottom beams is equal ; and the density of 

 the absorbing substance under test is then determined by 

 means of the cos 2 law (the intensity of the light passing 

 through crossed nicols being proportional to the square of 

 the cosine of the angle between their planes of polarization). 



The instrument as used as a spectrophotometer will be 

 found very completely described (together with an admirable 

 bibliography on the subject of Spectrophotometry) in a 

 paper by Messrs. C. E. Kenneth Mees and S. E. Sheppard 

 printed in the ' Photographic Journal ' for July 1904. 



I might, however, again mention here the fact that the 

 wave-length of the light under observation is read off direct 

 on the helical drum of the micrometer-screw, see fig. 1. 

 This is an arrangement which is found of great convenience 

 in practice. The wave-lengths Q can be relied on to an average 

 accuracy of well within 10 Angstrom units. 



The present paper is principally concerned 



(A) with the mode of correction of a serious error to which 



polarization spectrophotometers are liable; 



(B) with the use of the instrument as a polarimeter. 



(A) The error referred to is that due to the polarization 

 of the light on transmission through the dispersing prism, 



2 L2 



