LAMP------>FLAME -----> Prism (monocromator) --> Detector (photomultiplier)
The light emits a known amount of light
the flame vapourises the sample. in being vapourised the element gives off a certain colour (as you know from doing flame tests for elements)
The element also absorbs a certain colour (wavelength of light)
the light from the flame is then split into it's individual wavelengths with the prism (so they can isolate the particular one for the element they are looking for)
the detector picks up this wavelength and compares it's intensity to the known intensity of the original light source (the lamp)
the difference in recieved intensity to send intensity is directly prportional to the concentration of the element in the sample (the absorbance reading ) . using this data you can compare this sample to known concentrations tested the same way.. comparing their absorbance readings to the sample's absorbance readings lets you plot a graph and estimate the concentation of your unknown sample