54 



PROF. COPELAND AND DR L. BECKER ON THE 



determined are erroneous by the amount of curvature of the lines, they have to be 

 employed in the reductions of the star observations in order to reduce the lines of the 

 stellar spectrum to wave-lengths. 



In the night observations I measured at the beginning and end of each set the 

 prominent spark-lines in the part of the spectrum under observation, always turning the 

 screw in such a manner that the viewer moved opposite to the direction of gravity. Their 

 readings thus being known, I could then also pick them up while observing the stellar 

 lines without being obliged to turn the screw in the opposite direction. The first obser- 

 vations serve to correct the reduction-table, the second to determine the changes of 

 the zero point. Since the spectroscope is not rigid enough for taking several pointings 

 of one line, without observing each time the spark spectrum, I measured through one 

 region of the stellar spectrum without turning back. For this reason all the observations 

 of any one line are quite independent. Almost all the measures were taken in bright 

 field illumination. It is needless to say that the observations of most of the lines were 

 very difficult, but I have not the least doubt that those repeatedly observed refer to real 

 bright lines, and are not merely an effect of contrast produced by dark lines on the 

 continuous spectrum. There are 302 observations, belonging to 71 lines, made on 

 February 3, 4, 5, 10, and 11. Afterwards I made arrangements for photographing the 

 spectrum, but the sky did not clear up before my return to Edinburgh. 



The mean values of the wave-length A and their intensity I, (where 1 stands for faint, 

 6 for very bright) are : — 





Number 











Number 







1892. 



of 

 Observations, 



A 



I 





1892. 



of 

 Observations. 



A 



I 



Feb. 4, 5, 10. 



5 



657-0 



5 



Feb 



3, 4, 5. 



5 



560-0 



3 



„ 3, 4, 10. 



5 



640-5 



3 



5) 



3, 4, 5. 



6 



5570 



3 



„ 4, 5, 10. 



4 



632-5 



2 



)) 



3, 5. 



5 



552-4 



2 



„ 5, 10. 



2 



624 + 



1 



)) 



4. 



2 



5510 



1 



„ 4, 10. 



3 



620-3 



2 



JJ 



3,4. 



3 



548-8 



2 



„ 3, 5, 10. 



6 



615-1 



2 



)) 



3, 4, 5. 



6 



544-6 



3 



„ 3, 4, 5, 10. 



5 



609-9 



2 



)) 



4. 



1 



543 + 



1 



„ 3, 5, 10. 



7 



604-7 



2 



)J 



4,5. 



5 



540-7 



3 



„ 3, 4, 5, 10. 



7 



598-5 



3 



)> 



5. 



1 



539 + 



1 



„ 3, 4, 5, 10. 



8 



593-4 



3 



>) 



4,5. 



6 



537-4d 



3 



„ 3, 4, 5. 



13 



589-7 + 0-15 



4 



J) 



3, 4, 5. 



6 



533-0* 



4 



„ 3, 4, 5. 



4 



583-8 



1 



JJ 



3, 4, 5. 



5 



528-0 



3 



,, 3, 5. 



3 



5S0-5 



2 



)) 



3, 4, 5. 



6 



524-6 



2 



„ 4,5. 



3 



577-1 



2 



)) 



3, 5. 



3 



519-6 



2 



„ 4,5. 



3 



572-9 



3 



>J 



3, 4, 5, 11. 



8 



517-45 + 0-12f 



5 



„ 3,4,5. 



4 



568-7 



2 



)> 



3. 



1 



513 + 



1 



„ 3, 4, 5. 



4 



564-9 



3 



» 



3, 4, 11. 



4 



511-1 



2 



* Many close lines in this part of the spectrum, of which this is the most prominent. 



t On Feb. 10 it was recorded that the bright line 517-4 is very broad, and that the intensity fell off gradually 

 towards the red, while it was cut off abruptly at the more refrangible edge. 



