SPECTROSCOPIC PHOTOGRAPHY 



809 



The fact that a grating throws the light which falls on it into a number of orders 

 naturally results in a loss of light in any one spectrum, but this loss can be reduced 

 somewhat by shaping the point of the ruling diamond so that more light will be thrown 

 to one side than the other. When very high resolving power is desired, a grating can 

 sometimes be found in which most of the energy is thrown into the higher orders, and 

 it is desirable, if possible, to have most of the light in the orders on one side. In 

 general, if a grating shows high intensity in one order in a given direction, all other 

 orders lying in that same direction will tend to be strong. Thus a grating which is 

 found to give high intensity on one side in the second-order green (5500 A.) may be 

 expected to be bright also for the infrared near 11,000 A. in the first order, though 

 other factors may prevent this. 



In selecting a grating, it should be tested for brightness in the various orders and 

 for resolving power, line shape, amount of scattered light, and intensity of ghost lines. 

 Ghosts are false lines produced by regularities in some irregularity of ruling of the 

 grating. In a good grating the intensity of a ghost line should be less than 0. 1 per cent 

 that of the real line corresponding to it. Lyman ghosts are very objectionable, but 

 fortunately these have been successfully eliminated in the better ruling engines. 

 Rowland ghosts occur near the lines which produce them and hence can readily be 



Fig. 3. — Diagram illustrating the principle of the Littrow mounting 



identified. While these can usually be kept to less than Kooo the intensity of the 

 parent line, a grating which has Rowland ghosts of Koo the parent line may still be 

 usable for many purposes. Lyman ghosts can be tested for by illuminating the grating 

 strongly with collimated light from a slit, behind which is placed a powerful mercury 

 arc, and by looking into the grating near the central image or elsewhere where no 

 spectrum lines should be visible. If colored lines are seen, they may be Lyman 

 ghosts, and the grating should be carefully tested photographically. 



Gratings should be kept free from dust, which can be removed by gently stroking 

 with a soft camel's-hair brush. When dirty, gratings should be cleaned by light and 

 careful rubbing in the direction of the rulings with a piece of clean cotton moistened 

 with absolute alcohol and dipped in very fine precipitated chalk or tin oxide. Great 

 care must be taken not to scratch the grating. A portion of the unruled surface 

 should be rubbed gently before the rulings are touched to make sure that no grittj- 

 particles are on the cotton. Fingerprints should not be produced on the grating face, 

 and acid or other fumes which tarnish the surface should be carefully avoided. 



Mountings. — The device of autocoUimation developed by Littrow is widely used 

 with either prisms or plane gratings, when it is desired to obtain high linear dispersion 

 by the use of a camera lens of long focus. The principle of the method is illustrated in 

 Fig. 3. The diverging beam from the slit S is made parallel by the collimator C and 

 enters the dispersing system D, which in the case illustrated is a half prism with a 

 mirror coating on its back face. The light is reflected froin this mirror, passes back 



