No abstract

Batterman 1 has recently reported some x-ray measurements on powdered Fe and Cu in which he concludes that the number of 3d electrons in Fe is 6.2 ±0.4 measured relative to copper (assumed to have ten 3d electrons) and is 6.9 ±1.0 relative to a NaCl standard. Simultaneously with reporting this he has made available to us samples of Fe and Cu powder which he used in making these determinations and we are grateful for his consideration.We have repeated his measurements and made several of our own and while we find no major differences in the raw data, our analysis of the data fixes the number of 3d electrons in Fe at approximately 2.5 ±1.5. The differences in the analyses undoubtedly arise from factors such as absorption coefficients, Honl corrections, Debye-Waller factors, extinction corrections, and packing effects which cannot all be evaluated until Batterman reports these in detail. However, some of these differences can be discussed.Firstly, Batterman has used tabulated Honl corrections whereas we have measured these and find them to be -1.9 for Fe and -1.84 for Cu (Fe Ka). These have been measured by comparing the intensity ratio of the Fe 211 and Cu 311 peaks to an Al 311 peak at both Mo Ka and Fe Ka. In this comparison the Debye-Waller factors, preferred orientation, and any uncertainty in the scattering factors approximately cancel out and the only major difference in the ratios at the two wavelengths is the Honl correction at Fe Ka (the correction is small at Mo Ka). Secondly, we have found large packing density differences in