After the crystals had been soaking in their respective heavy atom solutions for
a week the prospective derivatives were screened by collecting diffraction data to a maximum
resolution of 
Å
on the X31 beam-line. The X-ray energy for each set was tuned
to be 
above the 
absorption edge of the heavy atom in question so as
not to necessitate the use of the calibrator for the screening process. In the majority of
cases this meant that a complete set of diffraction data could be obtained in less that four
hours depending on the angular setting of the crystals.
To test for the presence of heavy atoms, native data were collected to 
Å resolution
and used along with phases from a refined model of the hen egg-white lysozyme [8]
structure found in the Protein Data Bank at Brookhaven to calculate isomorphous and anomalous
difference Fourier maps (see Sec. 
).
The model was first subjected to eight cycles of least squares refinement against the native data to
ensure that no large differences existed between the two and the resulting phases used in the
Fourier synthesis. Data on the iridium derivative crystals were collected
after soaking for 10 days. The Bijvoet difference Fourier map calculated with data lying between
and Å is shown in Fig. . The map represents a projection
down the z-axis through 
of the unit cell, an asymmetric unit. There is evidence in the
map for the presence of anything up to ten anomalously scattering iridium sites in the asymmetric
unit, i.e. ten iridium atoms per lysozyme molecule. The isomorphous difference Fourier maps
calculated against the native data also supported the presence of multiple iridium sites. The
predicted Iridium site of Blake [11] was not present in the difference Fourier maps.
Figure: Difference Fourier synthesis calculated with the coefficients 
from the iridium
derivative data and phases from a native lysozyme model obtained from the Protein Data Bank. The map
for only th of the unit cell along the z direction is shown. Five strong peaks in the
map are clearly visible and several weaker peaks are also present. The map is contoured at 
intervals starting at 
and the largest peak reaches 
.
Anomalous and isomorphous difference Fouriers calculated for the other four potential
derivatives 
, 
, 
and 
displayed no evidence for bound
heavy atoms. In each case no significant peaks were observed in the Fourier synthesis above a 
cut off level.