Thermocouples ?
I reckon the sensor in question is most likely a
glass-bead thermistor - there are several types out there, both positive temp coefficient and negative temp coefficient (resistance rises with temperature increase and falls with temperature increase, respectively).
The resistance could be anywhere from around 50 ohms upto to a few hundred k, according to temperature and design.
Try a few resistors as substitutes, find a value that makes the indicator read zero degrees C and then a value that causes a reading of 100 deg. C. These values will help narrow down the choice of thermistor -- or prove me a complete waffling idiot
My only other guess would be that it is a 'Resistance Thermometer' - similar operation, same method of narrowing the selection of a replacement, different method of construction.
On a different note ...
Thermocouples
are subjects of the Seebeck effect. Trying to solder two wires together (introducing an intermediate metal) will spoil the thermocouple 'calibration' because of the tainting that takes place on and close to the surface of the two metals (they are no longer pure).
The Seebeck effect allows for "The Law of Intermediate Metals"; a third, fourth etc. metal as long as the additional junction temperatures are known and allowed for. In this case (a soldered joint) it can be argued that the new junctions (wire1-to-solder and soder-to wire2) are both at the same temperature and therefore cancel, giving no offset at all ! (there is still a pollution issue though).
To form a junction the two wires should ideally be welded, a good tight twist would suffice though :!: