The calibrator has allowed the instrument function and stability of the X31 beam line to be
characterised. The apparatus also proved useful in assessing the alignment of the focussing mirror
and has provided the option of recording absolutely calibrated 
and 
spectra. Exact knowledge of a beam line's energy resolution is important for several reasons. The
measurement of fluorescence spectra around the absorption edges of a heavy atom in the protein
environment is a necessary requirement for any optimised MAD experiment. As we have seen the form
of the scattering factor spectra is dependent on energy resolution of the beam line used to measure
it and it is therefore crucial that the energy resolution be quoted along with the spectra. With
anomalous scattering spectra measured on one beam line one could make an estimate of how the spectra
measured on the same sample would appear when recorded on another beam line assuming that the
instrument functions of both beam lines are known. This would mean that anomalous scattering spectra
could be considered transferable between beam lines having different characteristics. However this
also relies heavily on the energy calibration of the beam lines in question.
As was seen at the end of Chapter 
where data measured around the Cu
edge on two different beam lines are compared, the incorrect energy calibration can considerably
effect the values of the anomalous scattering factors at energies around the absorption edge. Recent
work by Tesmer et al [109] has appeared in which anomalous scattering factors were
published as a function of energy for the 
edge of mercury bound in a protein to a
cysteinyl sulphur and an aromatic carbon in p-chloromercuribenzylsulfonic acid (PCMBS). The
authors stated that the anomalous scattering factors should be applicable to other proteins with
mercury bound in this manner. However care should be taken when transferring such data for use on
another beam-line. Unless the energy calibration of both beam-lines is well understood (i.e. the
beam line where the anomalous scattering factors were initially measured and that where the MAD
experiment is to be performed) the transferability of such data may be inappropriate. The situation
will often be that fluorescence can be directly measured from the protein sample. If this is the
case then anomalous scattering factors can be freshly established before the MAD experiment
begins. If the samples are however too small or dilute in terms of heavy atom concentrations to
allow direct fluorescence measurements to be made then one must rely on spectra recorded from
samples which contain the heavy atom in a similar environment as that expected in the protein as
suggested by Tesmer et al. One possible way of making fluorescence data transferable between
beam lines is to calibrate the spectra with respect to some fixed absorption feature, e.g. the
absorption edge from the heavy atom in a standard sample of the metal oxide or alike. The
calibration apparatus described herein dispenses with the need for measuring standard reference
spectra as each individual spectrum can be placed on an absolute scale.
To summarize, the calibrator provides a monochromator independent and, if care is taken, a source independent measurement of the X-ray energy continuously during an experiment. Use of the calibrator ultimately saves time recalibrating the X-ray energy to the absorption edge of interest during an experiment or after a fresh injection of electrons and simultaneously minimise the amount of irradiation of the protein sample necessary for calibration purposes. It also allows absolute energy calibration of fluorescence spectra and can be used to determine the bandwidth of the X-ray instrument as well as to diagnose problems with mirrors and other optical elements. The problems associated with the use of the calibrator for energy stabilisation at other absorption edges may be overcome by minor adjustments in the design.