temperature coefficient for tuning forks

[Robert Scott]

Some brief experiments with a fork at 100 degrees F and 45 degrees F
leads me to believe that the temperature coefficient for the fork I
have is about -0.1 cents per degree F.  Can anyone confirm or refine
this number? Also, perhaps the metalurgists among you could comment 
on whether different forks might have different temperature coefficients,
and over what range can I treat the temperature effect as linear.  I
am planning on filing Ron Torella's forks down to 415 and 430 Hz and
I would like to know what corrections I should make for temperature.

Bob Scott
Ann Arbor, Michigan