NO. 8 WATER-VAPOR TRANSPARENCY FOVv^LE I9 



fall near the slit. Finer adjustments brought the image so that its 

 light passed centrally through the slit. The spectroscope collimator 

 and diaphragm were then adjusted so that the beam was central 

 both on them and on the prism. The last condition was best ascer- 

 tained by looking towards the collimator through the prism and 

 adjusting so that -the filament was visible over the whole extent of 

 the prism. The bolometer was adjusted to look centrally at the 

 projection of the beam on the image-forming mirror and the latter 

 adjusted to bring the image of the spectrum vertically on the 

 bolometer. 



Then, when using the 60° prism, sun-light was thrown on the 

 spectroscope slit and the prism ^ adjusted in minimum deviation 

 by slightly turning it back and forth, leaving it in the position where 

 the A line, as seen in the bolometer eye-piece, was at its minimum 

 deviation. With the 15° prism the purity of the spectrum was 

 insufficient to show any solar lines. A Bunsen flame, colored with 

 sodium, was used to illuminate the slit instead of sun-light and the 

 adjustments were made by means of the D lines. The spectrometer 

 circle was then turned to bring the A or D line central on the 

 bolometer strip and the position was read. Energy curves (or 

 holographs) were started from the spectrometer positions thus 

 determined. 



After the apparatus had been thus adjusted the blower was started 

 and readings were made of the wet and dry thermometers in the 

 water-vapor tube before beginning observations. 



OBSERVATIONS . 



It was originally proposed to observe alternately the energy in 

 the spectrum of the Nernst lamps when shining through the water 

 vapor in the large tube and immediately thereafter when shining- 

 through the spectroscope alone. It was found, however, that the 

 curves obtained through the water vapor in the spectroscope re- 

 mained so constant over the greater part of their wave-length range 

 that in the 1916 observations they were often omitted. The change 

 in aspect of the curves attending the above described change in 

 path of the rays gave a measure of the absorption produced by the 

 water vapor and carbon dioxide in the tube. 



^ Before setting- up the prism it was best to give it a new coat of asphalt 

 and slight!}' warm it out in the sun-light to prevent the fogging of its surfaces 

 by the deposition of moisture from the air. 



