I was able to rescue quite a nice thermometer from the surplus equipment left over when the calibration facility at my (former) workplace was closed down recently. It's a Labfacility Tempmaster 100, and it came with two PRT probes, one of which is an Isotech 935 14-13 'semi-standard' probe, rated from -196?C to 250?C, with typical uncertainty of around 10mK after calibration, and typical drift of 0.01?C/year at 0?C. Unfortunately, it doesn't have individual calibration data, but the specified tolerances on alpha and R0 should give an uncertainty of around 0.1?C over the 0?C to 100?C range, at least. An ice-point check with this probe read -0.02?C, which is very encouraging. The second, 'industrial' type probe read -0.04?C.
A random selection of K-type thermocouples and meters was anything reading from -0.8?C to 1.7?C in the ice-point bath, which sounds terrible but is within their specified tolerances! Precision temperature measurement - say to better than 0.5?C uncertainty - is hard.