Rollins — Riding Concave Gratings. 



49 



Art. II. — On Ruling Concave Gratings / by William 



KOLLIKS. 



Lyman has shown that the Rowland concave grating gives 

 false spectral lines so sharp and clear " there is probability and 

 some evidence " they have been mistaken for real lines. As 

 the concave grating is a beautifully direct method of obtaining 

 spectra, it seemed desirable to consider the causes of its defects 

 and try to remedy them, rather than to return to flat gratings 

 with their lens complications. 



Briefly stated, the process of preparing a concave grating is 

 to grind and polish the surface of the metal block to the 

 proper curvature, afterward mounting it on a carriage moving 

 on straight ways by means of the long nut of the precision 



screw, which carries the metal blank under the diamond as the 

 screw turns. This method, while well adapted to making flat 

 gratings, does not yield concave gratings in which the grooves 

 cut by the diamond are the same distance apart, and cannot 

 produce them with sides having equal faces or forming the 

 same angles with the surface of the metal in different parts of 

 the grating. The shorter the focus of the grating the more 

 marked are these defects. 



Figure 1 is a diagram intended to show them exaggerated. 

 The lines B to K represent equal movements of the grating 

 under the diamond point, as produced by the uniform motion 

 of the screw. If the grating were flat the resulting cuts would 

 be equally spaced, while the angles formed with the surface by 

 the sides of the cuts would be alike in all parts of the grating. 

 In the case of the curved grating the distance between the cuts, 

 measured on the surface of the grating, increases from the 



Am. Jour. Sci.— Fourth Series, Yol. XY, No. 85.— January, 1903. 

 4 



