Here is one possibility to consider. Add an oscilloscope across the
resistor and see if you have a lot of ripple on the DC. If the DC power
supply is not regulated, there could be significant ripple, and the
ripple will be worse at higher current.
If you have a lot of ripple your voltmeter and ammeter probably are not
reading the correct values. You need the rms values of the voltage and
current, but typical multimeters set to DC voltage and DC current will
not measure the true rms values of a fluctuating voltage and current.
Here is another possibility. Make sure your voltmeter is measuring the
voltage across just the resistor and not the voltage across the resistor
and ammeter. There is a voltage drop across the ammeter and thus some
heat dissipated there. If your voltmeter is reading the voltage across
the power supply, as opposed to the voltage across the resistor, you
will overestimate the power in the resistor, and it will be worse when
the current is higher.
You also have slight voltage drop in the lead wires, but that is minor
compared to the voltage drop across the ammeter. I think lead-wire
resistance is not likely your problem. However, to eliminate wire
errors, run a separate set of wires from the resistor to the voltmeter
so you are measuring just the voltage drop across the resistor, or
better... the voltage drop across everything that is under water.
Summary... If there is a lot of ripple, you probably have incorrect
values for both current and voltage. If there is not a lot of ripple,
but you are measuring the voltage at the power supply, you have an
incorrect voltage. And, of course, you could have both problems.
Michael D. Edmiston, Ph.D.
Professor of Physics and Chemistry
Bluffton University
Bluffton, OH 45817
(419)-358-3270
edmiston@bluffton.edu