100 DR L. BECKER ON THE SOLAR SPECTRUM. 



lathe suitably altered to replace the one first used. Since it formed a part of the record- 

 ing apparatus, it will be described later on. 



Dr Copeland was kind enough to undertake all the trouble of arranging and superin- 

 tending the erection of the observing station, where camping accommodation was also 

 provided on account of the short interval between sunset and sunrise in summer. The 

 first week Dr Copeland camped out, but had to abandon his intention of partaking in 

 the work, because the preparations for observing the solar eclipse of 1887 in Russia 

 required all his attention. There was only one fine sunrise during that week, when the 

 Rain-band was observed. But on account of the difference between our scales of inten- 

 sity, it was necessary to exclude these observations. For the same reason I also struck 

 out the first four sets of my observations which were taken in the first spectrum. The 

 whole of the later readings were made on a uniform plan in the second spectrum. 



During the years 1887 and 1888, I spent all the nights on the hill from the begin- 

 ning of June to the middle of August, by which time the lengthening nights made it 

 practicable to sleep below. This change was by no means' unwelcome, as the slight 

 wooden hut became very damp as the season advanced. The observations, however, 

 were continued up to the end of September. It would have been advantageous to have 

 extended them over the winter, but a range of hills to the south of the Baxmekin, which 

 rise about 4° above the horizon, prevent all observations at the most desirable altitudes, 

 while the prevailing mild winter temperature offered little change in the atmospheric 

 conditions. It was therefore considered not worth while to alter the position of the 

 heliostat, which was standing too near the hut for winter observations. 



The top of the Barmekin being 850 feet above the level of the sea, and about 14 miles 

 from the coast, is a most suitable place for this kind of work. To the east many of the 

 hills are below the horizontal plane, and the sea can be seen at various points ; to the 

 west there is a range of hills about 7 miles distant, rising at but a few points more than 

 a fraction of a degree above the horizon. Further south, however, as we have said, the 

 Deeside hills obstruct the view considerably. 



The original working plan of restricting our survey to the yellow part of the spectrum 

 was soon abandoned, and we proposed observing as much of the solar spectrum near the 

 horizon as possible. In 1887 the lines from A =6030 to b were surveyed, and in the 

 following year the tract from b to F was added. Beyond F it was found impracticable 

 to proceed, as the finest condition of the atmosphere, such as very rarely occurs, is required 

 to see the fainter lines when the sun approaches the horizon. There was already great 

 difficulty experienced in observing the spectrum from c to F, and we had often to break 

 off work at this part of the spectrum, because the wires of the viewing telescope were no 

 longer discernible, although there was still sufficient light to study the less refrangible 

 parts of the spectrum. Three sets of observations were added in 1889 to settle some 

 doubtful points. 



It may be mentioned that in the summer of 1 889 the work was continued below X = 6030 

 towards the red end, but this section was not completed when I removed to Edinburgh 



