Purpose: check whether even harmonics of current increase strongly with 
 the presence of a battery (even harmonics indicate a non-symmetric 
  waveform, whereas odd harmonics alone indicate a non-sinusoidal but 
  symmetric +/- waveform) 

A battery has presumably rather different processes happening when 
current is forced in the different directions ... indeed, our batteries
are all with 'primary cells', intended only for discharging, not
charging.  So it's very reasonable to consider a different resistance 
presented by the battery for the different directions. 

Our checks of linear capacitors without and with series batteries have 
shown no effect of battery polarity. However, the battery in these cases 
has pretty much an AC current (other than any conductance in the
capacitor, or still-happening polarization current, caused by the
battery's voltage) and the measurement of fundamental frequency values 
treats positive and negative directions equally, so even if the battery
were in fact behaving quite differently in the two directions, it would 
be expected to give the same result. 
Our check using a strong conductance so that there would be a significant 
dc bias in the current was only able to drive this dc component of current 
in one direction *relative to the battery* since it was the battery's 
own resistance through a positive resistance that drove it!

Why should we care anyway...? 
Mainly interest, for more confidence about what range of situations, currents, 
nonlinear test-objects etc, this method could be used for. 


