[keithwalton]

hmm i've never dealt with gas bottles myself but speaking as an engineer, i'll try and answer some of the stuff said above.

Gas bottles i presume are presurised with gas to weight. The pressure would be quite alot greater than atmospheric. It can also allow it to be storred as a liquid rather than a gas (LPG is like this i believe)

The gas will be let out in a controlled manner by using a throttle valve which will limit the pressure and flow rate of gas leaving the bottle to a safe amount. The pressure storred in the bottle will be much greater than this.

If the gas was storred as a liquid in the bottle (kinda ironic aint it) then it would settle at the bottom and have a finite level much like a bottle of water. You should then be able to read where this level is by some sort of permiativity device.
If it's storred as a gas then you'd be stumped to find a level as it would expand to fill its container. only weight would then do it.

Because of the throttle mechanism you will only be able to tell when the gas is about to run out a few moments before it does as the gas supplied will be a little over atmospheric pressure.
1.5 Bar as an estimate whilst internally it'll be say 15 bar initially.
Until it falls to 1.5 bar internally you wont notice any difference in performance then once it passes the threshold it will fall to what it is internally until it hits atmospheric pressure to when it'll stop coming out.

To summerise, better to be safe than to run out. If its storred as a liquid then you should be able to find it's level

K