276 PROCEEDINGS OF THE AMERICAN ACADEMY 



vision, 250 mms., or ten inches. Also call c the focal length of the 

 collimator, o that of the observing telescope, and e that of a lens 



equivalent to the eye-piece. Then Tr= - — ^^. Again, the dispersion 



ec 



or interval between the two images will equal that of the prism or 

 grating, multiplied by the magnifying power of the observing tele- 

 scope, or -, and will be quite independent of the collimator. 



e 



As in the microscope and telescope the highest powers are by no 

 means those which give the best results, so in the spectroscope the 

 best effects are not always obtained with tlie greatest dispersion. In- 

 creased angular dispersion is readily obtained by using a high power 

 with the observing telescope, but the limit is soon readied, since the 

 apparent width of the slit, and the various distortions, are increased in 

 the same ratio. Moreover, with a very great dispersion the light is so 

 far enfeebled that the spectrum becomes faint and the slit must be 

 opened wider. 



Tlie most satisfactory test of the efficiency of a spectroscope, as of 

 other optical instruments, is to examine some delicate object and com- 

 pare the appearance with that obtained with other similar instruments. 

 Formerly the D line was used for this purpose, which with any but 

 the smallest instruments is seen to be double. In the solar spectrum 

 it was soon found that there was a third line between these, and 

 afterwards several other lines were noticed. The observations of 

 Professor Cooke (Proc. Am. Acad. viii. 57), however, showed that 

 most of these lines were due to the aqueous vapor in the earth's 

 atmosphere, and that their visibility therefore de[)ended very largely 

 on the condition of the air. The E line is free from this objection ; 

 and, as it contains many more components, it furnishes a much more 

 complete test. The following table gives the appearance of the E line 

 as seen with various instruments. A dash denotes that the line 

 opposite which it is drawn was visible. When a double line is seen as 

 single, one of its components only is marked. Tlie lines given in the 

 map of Kirchhoff are shown in the column headed Kir. Those given 

 by Angstrom are, in like manner, marked Aug. To obtain the lines 

 seen with various instruments of the largest size, I asked several 

 friends to draw all the lines they could see with their spectroscopes ; 

 and I take this occasion to express my thanks to them for the 

 results. The column headed Sh. gives the results obtained by Mr. 

 .Sharpies, with a large spectroscope belonging to Dr. Gibbs, having 

 six 60° prisms, filled with chemically pure bisulphide of carbon. 



