MANNER OF EXPERIMENTING. 15 



at its refracting edge {d of figure 3) illuminated the very limit of the 

 field of view as seen through the chosen slot of the shutter (L of figure 6). 

 The position of the mark on the handle of the cell (F of figure 2), with 

 respect to the horizontal scale in the plane of the slit-jaws, was then read off. 

 If the cell were then moved, ever so little, in one direction the width of the 

 brightl}^ illuminated field could be seen to be less than the opening in the 

 shutter; whereas, if the cell were translated in the opposite sense no increase 

 in the width of the illuminated field occurred. At this opportunity, eye- 

 observations of the absorption between 0.400/j and 0.625/ji were always 

 made and the facts recorded. 



When the concentration of the liquid in the cell was much too great or 

 far too small this instrument had to be cleansed and filled with a solution of 

 more suitable absorbing power, obviously, but when the concentration was 

 not too remote from the best value the effective depth of the cell was varied 

 until the desired result was obtained. 



All three screws were raised and regulated in an obvious manner when 

 prisms of liquid having nowhere infinitesimal thickness were wanted. When 

 layers of liquid of uniform depth were studied a system much like that 

 shown in figure 3, but which had for bottom a plane-parallel plate of 

 quartz 2 mm. thick, was substituted for the quartz-wedge system. 



CALIBRATION OF THE CELL. 



The diedral angles formed by the cell were calculated from the dimen- 

 sions of the instrument, and also from measurements made with a spec- 

 trometer. 



EXPOSURES AND SPECTROGRAMS. 



The majority of the spectrograms consist of three distinct photographs 

 taken side by side and as close together as possible. {See the plates.) The 

 width of each photograph was practically the same as the width of the 

 opening in the shutter L. Numerous trials showed that this field of view 

 was completely filled with light, with no overlapping on the grating-side of 

 the opaque portions of the shutter, when the length of the slit was dia- 

 phragmed down to 10.5 mm. Consequently, the sht was limited to a length 

 of a very little more than this number and the cell was moved along exactly 

 10.5 mm. between the taking of two adjacent photographic strips on the 

 same film. By this means, the thickness of absorbing liquid through which 

 the light passed to the very edge of one photographic strip was equal to the 

 thickness subsequently traversed by the light which recorded itself at the 

 contiguous edge of the adjacent strip. Of course, the best appearing records 

 were obtained when the film holder, actuated by the rack-and-pinion system, 

 was moved, by an amount exactly equal to the width of the opening in the 

 shutter. A casual inspection of the positives reproduced in the appended 



