“…It does not touch the primary beam and therefore does not need to be included in the beamline analysis. Apart from the relative positions of mirror and monochromator, the present set-up is equivalent to those described by Yoda (1984), Furuno, Sasabe & Ikegami (1987), Hayashi, Hamada, Suehiro, Masaki, Ogawa & Miyaji, (1988) and Riekel & Suortti (1991) for rotating-anode sources and those described by Barrington Leigh & Rosenbaum (1974), Webb, Samson, Stroud, Gamble & Baldeschwiler (1977), Hendrix, Koch & Bordas (1979) and Amemiya, Wakabayashi, Hamanaka, Wakabayashi, Matsushita & Hashizume (1983) for synchrotron-radiation sources. We have chosen this set-up as it is the most common, even though it employs a perfect-crystal monochromator and that the resolution function due to this has a very small contribution from the wavelength spread.…”