Towards the end of a poor, grade 4, night, Uranus was tracked, using the SCUBA map16 method, from (az,el)=(196,56) at HST 08:19 to (228,43) at 09:59.

The mean antenna leg temperatures had bottomed out at 1.4degC between 04:00 and 07:00, and had risen to 2degs at 08:15 and 3.5degsC at 10:00. Still, this is relatively 'cool' and the potential for 'spikes' is high. A detail shown below from yesterday's inclinometry data -- when temperatures were rather similar -- indicates that measureable pointing excursions could be expected at azimuths between 214 and 224. Elevation excursions may be rather small (~1") and difficult to detect given the reported seeing during the observations of >1", but azimuth excursions might reach 4", being the sum of the F1 term * cos(elevation) + F2 * sin(elevation).

The map16 tracking data were troublesome to reduce before the v.noisy bolometer (g16) was flagged as being of 'BAD' quality and essentially turned off. In the past this has also caused problems for the algorithm that determines the centroid of each integration, but the problems should affect all integrations equally. The position of the centroid in (daz,del) is shown below :

• The tracking capability seems on average quite good, although the stability of the centroid deteriorates towards the end as the seeing rises to 2-3".

• However, the surprise is what appears to be a (-10",-10") pointing excursion in (azimuth, elevation) at about az=215 and a (+5",-10") excursion at az=221. These azimuths are at the edges of the expected feature displayed at top. The features are narrow - about 1deg wide or less - and their amplitudes are considerably larger than anything anticipated.

• As stated above the data are considerably affected by the effort to quash the spurious noise in the rebinned maps generated by the removal of the g16 signal, and concern about the ability of an algorithm to determine the centroid under such conditions is legitimate. However, the largest excursion seen in the above data (at az=214.9) is confirmed by independent reduction of that integration and the use of a cursor to estimate the centroid.

• Similar data -- maybe in better conditions -- should be obtained at this and other azimuths where excursions are expected. Alternately, perhaps a tracking experiment of several hours on Uranus as it traverses the southern sky would be revealing. Unfortunately, there are no large 'spikes' in the SE part of the sky. On the other hand, such a long tracking experiment may reveal further unexpected features . . .