wrote:
1. Mounted next to the track for a week or so at a time.
2. No power, it'll have to have batteries.
3. Other info: time.
It would be handy to be able to measure the passage of trains (in case
temperature suddenly increases when a train passes, then gradually
dissipates), but I'm not sure how easy/difficult that would be.
The logger would have to be small enough to be used in tunnels.

I have used a Fluke "Hydra" data logger in a lab situation. I was
capturing the data on a PC, but some models of the Hydra have internal
memory. They can be operated from batteries. They can measure voltage,
current, resistance, and temperature. I know it recorded the relative
time (seconds since powerup) with each measurement, and I _think_ it had
a time-of-day clock if you wanted absolute time. The models I'm familiar
with don't have a housing that would stand up to being outside; if I
wanted to use one in your application I would put it in a big Pelican
case, maybe. The logger itself was about 3"x8"x8" (about 8x16x16 cm)
but the batteries and case would easily double or triple that volume.

I have never used one of these, but I know they exist:
http://www.onsetcomp.com/Products/Pr...al_logger.html
It has a thermocouple input, a clock, and a voltage input that might be
able to record a dry contact closure to record the presence of trains.
Getting the contact closure is left as an exercise for the reader.

All of this does assume that you have permission to hook things up on
the tracks. Otherwise the railroad will probably react badly to
unidentified boxes with wires hooked up to their track.

Matt Roberds