Just to follow up a little on Matthew's answer: The Darwin width is the angular width over which a particular reflection will diffract. A rocking curve measurement usually leaves one crystal at a fixed angle and rotates the second crystal. For a perfectly collimated beam, the resulting intensity would be a convolution of the two Darwin widths. In addition, real x-ray sources have a finite angular spread of the incident beam, so that the rocking curve profile is further blurred. For bending magnet beamlines on older sources, the angular spread of the source can dominate the rocking curve. Many such sources use a collimating mirror before the monochromator in order to reduce the angular spread of the beam on the monochromator. In most cases, both contributions (natural Darwin width of the reflection and angular spread of the source) need to be included to get an accurate rocking curve. --Matt