and an Optical Method of Calibration. 421 



double-image prisms or extra collimators and prisms. The 

 visibility of this bounding line reduces the sensibility at least 

 one-half. In the arrangement of Oornu, and later of Kundt, 

 of applying a diaphragm to the objective, similar difficulties 

 are met with. The method first applied by Babinet, and used 

 by Wild, T rami in, and others, of superposing the interference- 

 bands from two sources of light after passing through <polari- 

 ziug-systems, in addition to the serious fatiguing effect on 

 the eye, does not allow of the sharp contrast necessary for the 

 highest sensibility. 



In the different methods adopted by Lummer and Brodhun 

 of causing the separating line to vanish we have one of the 

 conditions for the greatest sensibility of the eye. If the 

 " contrast principle " introduced by them be further utilized, 

 we have the highest sensibility yet attained. In direct com- 

 parisons of white light this latter principle has been success- 

 fully applied ; but in spectrophotometry the optical difficulties 

 in the way of its use have prevented the obtaining of corre- 

 sponding results. 



In the measurement of the relative intensities we have a 

 variety of methods, most of which involve serious cutting 

 down in the intensity of the original source ; so that for many 

 experiments the field is too weak for the eye to make accurate 

 comparisons. This is true of polarizing-systems and also of 

 sectorial systems, both when used as fixed diaphragms before 

 the objective and when made to rotate before the slit. The 

 use of absorbing-screens or of diverging optical systems in 

 spectrophotometry is impracticable. It has also been shown 

 by Lummer and Murphy that the direct use of the slit-width 

 cannot be relied upon in all parts of the spectrum when it 

 exceeds a certain value dependent upon the dispersive power 

 of the prism and the nature of the source. In the Lummer- 

 Brodhun spectrophotometer, as constructed by Schmidt and 

 Haensch, together with a direct-reading variable rotating 

 sector, after the principle of Talbot, for varying the intensity, 

 we have a means of comparison of the colours of the spectrum 

 to a degree of sensibility not hitherto attained in other forms 

 of instruments. This sensibility quite exceeds the constancy 

 of the ordinary sources of illumination, except of incandescent 

 filaments and pencils actuated by the same electric current. 

 The want of a source of sufficient intensity, constancy, and 

 uniformity throughout the spectrum is still for many com- 

 parisons, particularly in the study of absorption, a serious 

 hindrance in spectrophotometric measurements. 



The elaborateness and expense of such an outfit practically 

 excludes it from general use in colour measurements. Tlvi 



