Series and Shapes

I've been feeling kinda of disappointed in the same metal water battery hooked in series. You see even though there is electricity coming from the cell the voltage fluxuates from around -200 to +200 millivolts. The good news about this is that leakage current from the meter would not do this but the bad news is that it swaps polarity. You can't have one cell change polarity or it effects the output power because now that cell is working against the others. One would think putting several cups in series that I should be getting a higher voltage, but some cell will change polarity and the voltage will not be as high. I even tried my old theory of determining polarity with the trick of using different size plates, I use to think the little plate was negative and the big plate was positive. Well the different plate size idea doesn't work, they still swap polarity's when ever they wan't. I've have tried testing every cell individually but there is always a few rebel cells that will change polarity.

I guess a little bit of better news is that some shapes work better than just a plain plate. Making a U shape out of the aluminum wire with a straight piece in the water made a good bit of power. But the best shape was a coil of aluminum wire with a straight piece in the water, that made over 600mV, thats a huge increase of the regular 100mV normal cell. I've rarely seen a cell go that high, but of source it doesn't stay there like all cell is likes to swap polarity too. Why does it like to swap polarity! What on earth would do this, Its almost like its making its own very low frequency AC current. Maybe a bridge rectifier on each cell might be needed.