220 HEMOGLOBIN. 



efficient by the long-continued labours of its author, differs from Yierordt's in 

 the mode by which the equalisation of the intensity of two beams of light is 

 brought about, the difference in mode requiring a spectrophotometer which 

 differs in important respects from the instrument already described. 



In Hufner's spectrophotometer there is a single slit, the ividth of which, 

 after it has been once adjusted, is never varied. 



The light which reaches one-half of this slit has been polarised by a small 

 JSTichol's prism (the polariser), whilst that which reaches the other half (which 

 in the determination of the value of e passes through the thicker stratum of 

 coloured liquid) is unpolarised. When these two beams of light fall upon the 

 refracting prism of the spectroscope, they are refracted and furnish two super- 

 posed spectra, of which that corresponding to the polarised beam is naturally 

 much less intense than the other. Before making any observations of e, the 

 two spectra must be equalised, this bein<* done by interposing a wedge of 

 smoke-tinted glass in the path of the unpolarised beam. Equality of both 

 spectra having been obtained, if a coloured medium be placed in the path of 

 the unpolarised beam, its spectrum will be correspondingly reduced. Equality 

 is, however, restored by rotating a second Xichol's prism (the analyser) which 

 is in the path of the beams issuing from the refracting prism, and the rotation 

 of which diminishes the intensity of the polarised beam alone. When equality 

 in the illumination of both spectra has been restored, the angle (</>), through 

 which the analysing Nichol has been rotated, is measured in two opposed 

 quadrants of a divided circle provided with a vernier, and from the value of 

 < that of /' is calculated. 



If the original intensity of the light = 1, and the intensity of the un- 

 absorbed light which has traversed the coloured medium be represented by 

 /', then 



/' = cos 2 < ; 



If the layer of coloured liquid investigated be always = 1 (e.g. 1 cm.), then 



as e = log /', 

 e log cos 2 < 



The following example will illustrate the mode of procedure and the steps 

 of the calculation in an actual experiment for the determination of the ex- 

 tinction-coefficient of blood, carried out with Hiifner's spectrophotometer : 



1 c.c. of defibrinated blood of the ox was diluted to 160 c.c. with a O'l 

 per cent, aqueous solution of Na(OH). The absorption-trough was filled with 

 some of the perfectly clear red liquid thus obtained. The spectral region (r), 

 for which e was determined, was one of the two in which Hufner has, in his 

 most recent experiments, determined the constant A of oxyh?emoglobin (i.e. 

 a portion of the region between the bands a and /3 of oxy haemoglobin). 



r=A557'5~A568-7 



(Mean of ten measurements) <f> = 61 -8 7 

 Converting the decimal 



fractions of a degree into 



seconds < =6152' 



It has been stated that with Hiifner's spectrophotometer 



/' = COS 2 c 



and e - log cos 2 < 



In the above experiment 



e= - log cos 2 6152' 

 "-2 Iogcos61 52' 

 " = -2 (0-67350-1) 

 " = -1-34700 + 2 

 " = 0-653 



