The Ultra-Violet Absorption Spectra of Blood Sera. 



329 



empty observation cell in each ; or if a solution is to be examined, cells 

 of the solvent are interposed. The two spectra should be identical, and 

 thus show that the instruments are in proper adjustment. This photograph 

 is conveniently called a " test band." 



One of the empty cells is then replaced by a similar cell containing a 

 layer of the serum, the thickness of which has been determined as accurately 

 as possible, say to the one-thousandth part of a millimetre. This is a 

 matter of great importance, as the layer is usually less than one-fifth of a 

 millimetre thick. Thus a possible error of one-half per cent, is assumed, 

 and this postulates a similar error in measuring the magnitude of the 

 absorption. 



The photographic plate is lowered in the camera to expose a fresh strip 

 of the plate. The sector in the other path of light is set at the aperture 

 0'9, by which the intensity of the light is reduced to - 9 of the original 

 intensity, and a fresh photograph is taken. The plate is again lowered, 

 the sector set at the aperture 0*8, and another photograph produced. The 

 process is repeated with as many more sectors as desired, usually not less 

 than eighteen and not more than fifty-four. The plates adapted to the 

 spectrograph camera measure 10 inches by 4 inches and accommodate 

 18 photographs. Finally another test band is registered to show that the 

 adjustment of the instrument remains unaltered. 



The plate having been developed and dried, the points of equal intensity 

 in each pair of spectra are marked by small dots. The wave-length of each 

 of these points is determined by reference to a corrected wave-length scale 

 photographed on the plate, and tabulated along with the corresponding 

 sector aperture, that is, along with the corresponding intensity of the light 

 transmitted. In accordance with the usual practice this is applied to 

 determine the absorptive power of the substance in terms of the quantity 

 log (I/I'), known as the " extinction coefficient," where I = the initial 

 intensity, and I' = the intensity of the unabsorbed light. 



In view of serum being a solution, the composition of which is variable 

 and comparatively unknown, the extinction coefficient must apply to the 

 serum in some form arbitrarily chosen as a standard. The standard adopted 

 is a layer of serum one-tenth of a millimetre thick. When the substance is 

 employed in a thickness or concentration other than that prescribed for the 

 standard, the extinction coefficient must be calculated according to the 

 formula cd/c'd' . log (I/I'), where d is the standard thickness, c the standard 

 concentration, d' the thickness employed, c' the concentration employed. 



Since existing work on the spectra of blood is recorded in wave-lengths, 

 it is desirable to continue this in order to correlate the new work with that 



2 E 2 



