DR L. BECKER ON THE SOLAR SPECTRUM. 105 



4. The Observations. 



The working plan of 1887 embraced the region 603 to b, to be observed both in the 

 evening and morning in both the second spectra. For the sunsets the plan was carried 

 out, whilst the sunrise observations are complete in only one of the spectra. In the 

 following year, 1888, the deficiency of observations in the one spectrum was made up to 

 such an extent as was consistent with the idea of finishing the spectrum up to F during 

 that year. The sun at medium altitudes has been observed at least once on both sides of 

 the normal in the two spectra. 



On beginning a set of observations the first line was identified in Professor Smyth's 

 maps of the Solar Spectrum. This work has afforded us the greatest help, not only 

 while the observations were in progress, but also in the identification of the standard 

 lines in the reductions. 



The day of the month was entered on each strip, followed by readings of the dry and 

 wet bulb thermometers and notes about the weather. Sometimes the intensity of one 

 or two, water- vapour lines between the sodium lines was also noted. The time at which 

 a line was observed was occasionally put down, to allow of computing the sun's altitude 

 above the horizon and the quantity of air traversed by the rays that entered the spectro- 

 scope. At the end of each set the last line was again identified and the meteorological 

 observations repeated. 



As the five needles of the recording apparatus made marks in a line at right angles 

 to the motion of the paper, any of them sufficed to record the position of the grating ; 

 while certain combinations served to describe the lines. Of the 19 signals employed 

 14 refer to the comparative blackness of the lines. With the left hand resting on the 

 five levers the thumb was used to indicate 1, and the consecutive fingers respectively 

 2, 3, 4, 5. This covers the five lowest classes of intensity. The following numbers up 

 to 14 were obtained by addition, thus: — 



6 = 5,1; 7 = 5, 2; 8 = 5, 3; 9 = 5,4; 10 = 5,4,1; 11 = 5,4,2; 12 = 5,4,3; 13 = 5,4,3,1; 



14 = 5,4,3,2. 



Since the main, object of our work was the determination of telluric lines, it was of 

 the utmost importance to retain a uniform scale of intensity that did not alter during 

 the time the same part of the spectrum was under observation. All the observations 

 had therefore to be made in either of the two second spectra and with the same eye-piece 

 on the viewing telescope. We defined by intensity = 3 the faintest lines that could 

 just be distinctly seen across the whole breadth of the second spectrum, while those 

 barely visible were designated by 1. The two webs forming the cross in the field of 

 view being of different thicknesses were used as standards for intensities 6 and 8. The 

 remaining classes were correspondingly estimated. If a line appeared much broader than 

 corresponded to its intensity, the record of the line was closely followed by a sign made 

 with the needles 3 and 2 ; and when the breadth was considerable, both its borders were 



