A method has been tried for obtaining accurate values of lattice spacings from X-ray powder photographs taken in the usual circular type of camera. There are two essential features of the method, (a) the calibration process, (b) the extrapolation process.
The calibration process: The exposed portion of the film is limited by sharp knife edges, and the length Sk is measured for each film at the same time as the distance S between corresponding pairs of lines. Sk corresponds to an angle θk in the same way that S corresponds to a glancing angle θ, so that θk = Sk / 4R, where R is the radius of curvature of the film. R is an uncertain quantity depending on the amount of film-shrinkage, and so in calculating θ from S we replace R by θk, which is a constant for a particular camera. θk having been determined in a preliminary experiment, it is possible to calculate θ for a given reflection, from the formula
θ = (S/Sk) θk
Errors due to film-shrinkage are hereby eliminated.
The extrapolation process: Other errors due to absorption by the specimen and eccentricity of the specimen may be eliminated by plotting the values of the lattice spacing calculated from a given pair of lines, against the corresponding values of cos2 θ. For small values of cos2 θ, the curve is almost linear, and is easily extrapolated to cos2 θ = 0, where the correct value of the lattice spacing is to be found. Only a few accurate measurements at angles where cos2 θ is small are necessary in order to carry out this process.
As examples of the method, results are given for the lattice spacing of iron taken in different cameras with specimens of different diameters. They are consistent to 1 part in 15,000, the mean value being 2.8605. A specimen of electrolytic nickel (3.5162 Å.) was found to give a different value from a specimen of Mond nickel which had been degassed (3.5170 Å.).