Magnetic field from a Zeeman diagram ($B=3s_E/(2{\mu }_B)$)

For weak-field sodium D lines neglecting hyperfine structure, the smallest adjacent spectral spacing in the full electronic pattern is $s_E=2{\mu }_{B}B/3$. A drawing without a labelled energy axis can still estimate $B$ if it includes both unsplit D lines with known wavelengths: their separation calibrates $hc(1/{\lambda }_2-1/{\lambda }_1)$. Divide the drawn split spacing by the drawn unsplit separation before using that energy scale. Graphical line thickness, unequal drawn spacings and uncalibrated rendering limit precision; they do not justify exact numerical field values.