ON THE PROLONGATION OF SPECTRAL LINES. 

 By Theodore Lyman. 



Received May 27, 1903. 



Investigators who have worked with concave diffraction gratings 

 cannot have failed to observe the faint but sharp prolongations of strong 

 lines which occur in the spectra produced by these instruments. 



The cause of the difference in length of certain lines when a prism 

 spectroscope is used is well known. Sir Norman Lockyer long ago 

 made use of the phenomenon of " long and short lines " in his study 

 of the chemistry of the Sun. He pointed out that when the vertical slit 

 of the collimator is illuminated by the image of a light source formed by 

 a lens, some of the lines in the resulting spectrum are longer than others. 

 This of course arises from the fact that the portion of the source which 

 illuminates the centre of the slit possesses some vibration frequencies 

 which are wanting in those portions which come to focus at the top and 

 bottom of the slit. When a concave grating is used, the astigmatism 

 renders this phenomenon less striking. 



The prolongations of strong lines which are referred to above present 

 a different appearance, however, from those obtained with a prism spec- 

 troscope. They are, in fact, due to a different cause. As the author 

 had never seen any explanation of the matter, it seemed that an investi- 

 gation on the subject might prove of interest. 



In making some adjustment of a concave grating of twenty-one foot 

 radius the principal image of the slit came into view. It was at once 

 noticed that this image was prolonged into two narrow streamers, one 

 vertical and the other horizontal, each quite distinct and sharp. This 

 suggested that the vertical continuation of the slit image and the vertical 

 continuation of strong spectral lines were due to the same cause. Ob- 

 servations were accordingly made upon the principal image of the source. 

 In order, however, to simplify the diffraction phenomena as much as 

 possible, a circular opening of about 0.01 cm. diameter was substituted 

 for the slit. In order to reduce spherical aberration to a minimum the 

 light fell upon the grating at nearly normal incidence. In the final 

 VOL. xxxxi. — .3 



