Part of the secondary radiation belongs to the element
When primary X-rays strike matter, the matter sends out secondary X-rays. Barkla found that one part behaved like scattered incident radiation. Another part had a penetrating power characteristic of the irradiated element. Changing the radiator and comparing absorption counted against the idea that all secondary rays were simply redirected primary rays.
Separating the two parts required controlled comparisons of excitation and attenuation. A strong detector signal could not pick out the characteristic part by itself, because transmitted, scattered and newly emitted radiation could all contribute to it.
Sources: [1]
K and L were first sorted by experiment
The K and L labels marked characteristic families that differed in penetrating power and in what it took to excite them. Identifying them early on did not require a crystal spectrometer that resolved every wavelength. The shell interpretation came later and explained families that experiment had already separated.
The 1917 prize was for the characteristic radiation of the elements. Moseley’s work on atomic number and Manne Siegbahn’s later precision spectroscopy answered other parts of the measurement problem, and they should not be folded into one discovery.
Absorption can tell radiations apart without a full spectrum
For a narrow beam containing a single spectral component and a uniform filter, the logarithm of the transmission ratio gives an attenuation coefficient. The thickness must be known, scattered radiation excluded and detector response held fixed. A mixture of components gives a sum of exponentials, so one coefficient fitted to the mixture need not belong to any single line.
The modern expression clarifies the logic of a filter comparison. It does not assign modern photon energies to Barkla’s measurements. What the lecture says about radiation and atoms has to be read in the experimental context of its time.
Modern instruments record different quantities
An EDS detector sorts photon pulses by the energy they deposit. WDS uses a dispersing crystal to select wavelengths. Both can study characteristic X-ray emission, but their response functions and line overlaps differ from those of a filter measurement. XPS measures emitted electrons instead and needs its own kinetic-energy reference.
Once excitation, overlap and background have been checked, an identified line supports assigning an element. An uncorrected peak area is still not a concentration, and it does not uniquely determine a chemical state.
What to look for in the lecture
Read the award presentation printed before Barkla’s lecture, especially its absorption comparisons and K/L classification on pp. 389–391. Ask how the evidence changes when the radiator, incident hardness or absorber changes. The later lecture contains tentative claims, including a proposed J series, that are not adopted here. Then turn to Manne Siegbahn’s diagrams to see how resolved line positions test atomic energy differences.