Quote Originally Posted by Bruce Appel
Another vote for a Kodak proccess thermometer. I went through the same thing a few months back, got one cheap on ebay and have been delighted. In absolute terms I don't know how accurate it is, but it is repeatable, easy to read, built like a tank.
The only Kodak Process Thermometers that were calibrated and NIST certified were the "Type 3". They come supported in a stainless steel exo-skeleton for want of a better term.
BUT.... how do you still know that it is in calibration? Without a recent NIST traceable calibration you really don't know it is in calibration. Are you wondering whether an LIG thermometer can go out of cal? It sure can, check out this useful page of information as to how they can and do go out of cal.

http://www.icllabs.com/whyrecalibrate.html

Most liquid in glass (LIG) thermometers are 'repeatable' as you say (the more correct term is 'precise'). An LIG thermometer will most likely have high precision, but unless it has been recently calibrated with an NIST or equivalent certificate to prove it, then the accuracy cannot be guaranteed.

Unfortunately in light of this, the OP (Brian) is still left scratching his head, because he probably has no convenient way of ensuring that any LIG thermometer he buys is reading accurately. And unless he has a recent cal. certificate, unfortunately he can be none the wiser.

If you go for a digital one, how do you know that the Chinese brand $5 one is accurate to its stated specifications? The ones I would tend to trust more are the digital oral thermometers which unfortunately don't go down below 32.0degC.
Of all the digital thermometers mentioned in this thread, only one mentions an accuracy claim (by design) of +/- 1degC (http://www.rhdesigns.co.uk/darkroom/...ermometer.html). So just because your new fangled digital thermometer has 0.1 degC resolution, does NOT mean it is accurate to +/- 0.1 degC !! It is most likely accurate to something between +/-1 to +/-2 degC. Really !!!

I also wanted to highlight two other statements about digital thermometers in this thread that are rather misleading:

1. "I found a digital lab thermometer on Ebay a while back that seems pretty accurate"
according to who or what !

2. "I've been using a pyrex digital thermometer from Linens and Things for about 2 years and it's dead on with my glass one every time. Accurate w/in a tenth and it's quick as well"
just because it has a 0.1 degC resolution does NOT imply it has 0.1 degC accuracy ! Secondly, how do you know that your glass one is dead on with a traceable standard ???

Never be fooled into thinking that an LIG or digital thermometer's scale divisions are in any way a statement of its accuracy.

regards
Peter