## Abstract

X-ray diffraction effects in the intermediate plagioclase felspars An$_{1-x}$Ab$_x$, where $x ~\simeq \frac{2}{9}$ to $\frac{7}{9}$, are explained quantitatively, as regards breadths and intensities as well as positions of the maxima, by an ideal structure subject to stacking faults. This is deduced directly, by classical diffraction theory, from the published evidence about diffraction effects, without making any assumptions about detailed atomic arrangements. The unit cell of the ideal plagioclase structure has edges $\mathbf{a}_0$, 9$\mathbf{b}_0$, 2$\mathbf{c}_0$, where $\mathbf{a}_0 = \mathbf{a}_{Ab}, \mathbf{b}_0 = \frac{1}{2}(\mathbf{a}_{Ab} + \mathbf{b}_{Ab}), \mathbf{c}_0 = \frac{3}{2}\mathbf{a}_{Ab} + \frac{1}{2}\mathbf{b}_{Ab} + \mathbf{c}_{Ab}$. Sixteen of the eighteen subcells resemble closely one or other of the four different subcells of anorthite, and are juxtaposed so as to build up extensive anorthite-like slabs within the structure; the remaining two subcells form transitional sheets parallel to (010)$_0$. Regularly distributed stacking faults on (100)$_0$, with slip vector -2$\mathbf{b}_0$, are responsible for the most conspicuous variation with composition of the positions of `e' maxima (`split-b reflexions') and `f' maxima (`satellite spots'); stacking faults on (010)$_0$ and (001)$_0$ account for other features.