As mentioned I use a current limited supply for the production of rainwater CS for garden, and this is experimentation making a 100 gallon batch. The test at start showed 10 VAC variac input to the 30 ma current limited NST allowed 1.9 ma, with 7.9 DCV input becoming 7.5 DCV after 20 hours of current delivery using 10 ma conduction through the solution @ ~ 33DCV, where this voltage did not vary downwards over time as should be suspected. Then the one hundred gallons was mixed by drainage and repumping with another 30 gallon drum with mosquito larvae. First a sample of the mosquito water was taken from the drum in a gallon pitcher and then mixed by two/ thirds CS product from the tank. In two days the larvae have died in this sampling. The re-mixture of the CS tank water with one third volume extra water also showed itself as a higher resistance value where the new voltage across the tank with a 10 VAC variac input was 7.7 DCV. With the pump replenishing the volume of water expelled on the drain, the water is being continually remagnetized, and at this point miracle grow fertilizer is added in the mixing process. Now the resistance of the liquid water medium shows a ten fold reduction,( by electrolyte addtion) It now reads about .7 DCV from an open circuit ~1200 VAC source. I let it run about ten minutes in this recirculation of fertilizer water, then DC voltage went down to.6DCV input. Now I shut the power supply off and also pull the plug on the NST primary, and its DC conversion does not employ a DC filter capacitor that could hold charge. Again I read the ambient voltage of 1/3 volt. Now we start draining the tank as before. AS soon as the tank has drained to where the top electrode should not touch the water, as the top electrode is fixed in elevation and not a floating arrangement: instead of the voltage giong to zero, it reads a higher value.