If you're even slightly suspicious of the nicad pack then I would measure the voltage across each of the cells when not under charge, and again when under charge.
I would also measure the voltage across the charge current limiting resistor too, and this will determine the currtent actually flowing into the pack.
Each cell should measure about the same voltage.
Sure signs of problems are:
Very small (or even reversed) voltage on one or more cells when not under charge
Unusually large voltage when under charge.
If voltages appear OK, but are uneven, it may be a good idea to discharge the cells individualy until they hit 1V, and then recharge the pack. Hopefully they will all come up to about the same voltage.
Typically only a single cell in a pack fails, but you should always replace ALL cells.
To determine the charge current, use the formula i = v/r where i is in amps, v in volts (voltage drop across the charge resistor) and R in ohms.
grey red gold gold is 8.2 ohms by my calculation, but you may wish to check it (or my memory) by directly measuring it.
at this value, with a 9V charger (under load), and 3 nicads, you'd expect a voltage drop of around 5 volts and therefore a current of 600 mA and a power dissapation of 3 watts!
This is completely absurd, so clearly my calculation of the resistor value is way off (could it be 82 ohms grey-red-black-gold?) or the loaded voltage of the charger is lower (5v?).
But the point of this is that the charge current is about 0.1C (i.e. 60mA for 600mA/hr nicads). Any significant deviation above or below this may indicate problems.
If the current is low, and the nicad pack voltage is unusually high it may indicate that one or more of the cells is not accepting a charge.
If the current is high, and you tend to leave the charger on for extended periods, then one day you'll find the previous problem will occur :-)