29 pointing data with RxA/ACSIS were collected in grade 3 weather during a dedicated pointing run lasting about 3 hours. Removing collimation offsets of (CA,IE) = (-0.5",+6.9") gives the residuals shown below:

Raw rms residuals in (dS,dZ) are (1.7",1.9"). There is only one weak systematic trends remaining - that of dS -vs- Z (centre-left panel) - but otherwise this is a satisfactory confirmation of the utility of the model installed on 28 Feb. The concerns raised recently during the analysis of ~2weeks worth of serendipitous data must be explained away by nightly peculiarities.

A full model fit - to account for the 'new' systematic - reduces (dS,dZ) to (1.5",1.8"), and this model has been installed today.

11 data taken later in the night using the HARP-array show collimation offsets of (+16.6,+37.4") and have rms residuals (dS,dZ)=(2.6",3.6"). However, most of the error appears in one measure (#67 on NGC6334I) and without that one (dS,dZ)=(1.5",2.5").