Cu/Zn in salt water voltaic cell peculiar behavior -- any explanation? I was playing with my eight-year old son and his new 'lemon battery' kit -- it has 4 Copper plates, 4 zinc plates, a red LED, and pieces of wire hooked together. 
Having no lemons handy, I decided to use salt water (regular Iodized kitchen salt dissolved in tap water). Being lazy too, I decided 3 cells in series would be good enough for the LED to glow (I guess the $V_L$ of the red LED is ~2V ... and each Cu/Zn cell makes maybe ~1V each?) ... and sure enough, it is: the LED lights up, and my kids are suitably impressed.
Now the funny part -- since I made 3 cells, I still had one spare Cu plate and one spare Zn plate. 


*

*When I take the spare Zn plate, and touch the submerged part of the Cu (+) electrode of any of the three cells with it, the LED stops glowing.

*When I take the Zn plate away -- the LED starts glowing again.


So my question is, why is this happening? I could not answer my son when he asked: that was embarrassing! The LED does not stop glowing when I touch:


*

*A (spare) Cu plate to any submerged Zn (-) electrode 

*A (spare, dry) Zn plate to the non-submerged part of the Cu electrode  

*A (spare) Zn plate to any submerged Zn electrode  

*A (spare) Cu plate to any submerged Cu electrode


Another observation is, when I take the Zn plate away from the submerged Cu electrode, the LED lights up but doesn't reach full brightness instantly -- the brightness increases from zero to full over a period of about one second. 
Any guidance? Thanks...
 A: When you submerge either the zinc plate or the copper plate and touch the two together you are creating another (local) cell within your cell which is analogous to connecting a battery in parallel to one of three batteries connected in series (see illustration below). But when you do this you need to take care in how you connect that parallel battery.

If the the parallel (local) cell polarity is in the same direction as the main cell (A.) - then no problem you have the same voltage you started with. But if you connect the cell in the opposite direction (B.) you create a short circuit between the local and main cell. All the current flows in this short, and the current in the three cells in series goes to zero.

Connecting the extra copper to the zinc plate is like having a battery in parallel with the same polarity. And when the zinc plate is connected to a copper - opposite polarity. In your experiment, as in the case of connecting a battery in parallel, but 'backwards', when you connect copper to zinc nearly all the current you generate for that cell all circulates through the local cell - and the LED gets very little if any at all.
And the reason the LED takes time to illuminate is that unlike electrons in the metals, the ions that flow in the electrolyte don't (almost) immediately flow. It takes considerably more time for the ionic bridge to form.
