大家再请看下面的内容，我提此问题来源于此：

Wavelength calculation

--------------------------------------------------------------------------------

In the original data files from Kurucz only one wavelength value is listed for each line. For wavelengths below 200 nm this is the vacuum wavelength of that line; for those above 200 nm, it is the air wavelength, calculated from Edlen's formula (B. Edlen, Metrologia Vol. 2, 71 (1966)). (I'm not perfectly sure about this last point, that means with which formula Bob Kurucz calculated the air wavelengths.)

In this version of the spectral line database, searching for wavelengths is done on Kurucz' original data. That means, for wavelengths below 200 nm, the search is done for vacuum wavelength; for wavelengths above 200 nm, the search is done on air wavelengths.

In the one-column wavelength output, wavelengths are given as they appear in the original Kurucz data files, that is, vacuum wavelength below 200 nm and air wavelength above 200 nm.

In the two-column wavelength output from this search program, for wavelengths between (currently) 185 nm and 200 nm the air wavelength is calculated from the given vacuum wavelength using Edlen's formula.

Above 200 nm, in the two-column wavelength output the vacuum wavelength for a line is calculated by applying Edlen's formula to the air wavelength given in the data. However, because in this approach I had to insert the air wavelength into the dispersion formula instead of the vacuum wavelength, as it would be correct, in some lines with wavelengths above 200 nm a round-off error in the last digit of the vacuum wavelength value might appear.