Please note: the following article documents some experimentation on a novel technique for cycling a tank and filtration system without fish. This is not a foolproof method and requires familiarity with the aquarium nitrogen cycle and proficiency with the full set of test kits. If you are having trouble getting a tank to cycle with fish in it, this method offers nothing for you! Stick to the basics, changing water and limiting feeding until the bacteria finally establish. Think about the fish first ... and play around with techniques like this later when things are under control.
The process of developing the proper bacterial cultures in your biofilter to convert fish wastes (primarily ammonia) into less toxic compounds (primarily nitrates) is mentioned all over this web site. The reason it gets so much attention is that good biofiltration is absolutely critical to the health of our fish, but getting the bacterial cultures up and going can be a pain. It often can take two months for the ammonia and nitrite levels to finally drop to 0; two months of 50% per day water changes and stressed-out goldfish. Wouldn't it be great if both we and our fish didn't need to go through this?
In theory, it should be possible to cycle a tank without fish in it, by adding ammonia from a chemical source, at a rate similar to what you would expect from the anticipated fish load. In addition, we could raise the temperature and use bacterial starter products (like Fritz Zyme) to further speed the process.
James S. Koga published an article in "Freshwater and Marine Aquarium" (December 1996, page 213) on doing precisely this with household ammonia. His key points are that bacterial growth is optimal at 86 - 95 degrees Fahrenheit, that adding 3 to 5 drops of ammonia per day to the tank will provide sufficient food for the bacteria, and that a tank should fully cycle in two to four weeks.
Separately, in the spring of 1996 I had tried experimentally cycling a filter on a ten gallon tank with household ammonia. Adding Fritz Zyme, in two doses (an initial dose and then a second after there was measurable nitrite), plus ammonia daily, allowed me to cycle a filter fully in 9 days. At the end of that interval, the filter was able to process a quarter teaspoon of household ammonia per day which is roughly the bioload I had calculated for a large, well fed goldfish.
I felt pretty happy after that experiment ... I had the method down! Not so; I was unable to replicate that success. Why? Basic chemistry, literally!
Household ammonia is primarily a dilute solution of ammonium hydroxide. As such it is a strong base. Added to tap water, particularly a weakly buffered tap water like mine, will sky-rocket the pH level up past the eights, nines, and beyond. It turns out the ammonia oxidizing bacteria (that produce nitrites) can actually handle this to some extent. But the nitrite oxidizers (that produce nitrates) are more finicky. I was dumping expensive Fritz Zyme into an extremely basic solution, killing the bacteria I wanted, and then ignorantly blaming Fritz for making an inferior product.
(Aside: one general moral is that, when things don't seem to be working the way you expect in your aquariums, go back to the basics. Test your water parameters!)
I redid the "Fast Cycle" experiment, using white vinegar to keep the pH in bounds. And it worked! But it was a pain; the pH adjustments were very sensitive and I needed to add more vinegar twice a day. Apparently there are bacteria that breakdown the vinegar even faster than the ammonia. I had to do about 10 pH tests per day, to get it right (measuring the amounts of ammonia and vinegar didn't seem to be enough).
One way around this would be to try ammonia sources that would not have as dramatic an effect on the pH. Perhaps ammonium chloride (from an aquarium supplies store) or ammonium sulfate (from a garden center ... but careful about contaminants!). Or just use a few drops of household ammonia per day, check the pH, and do water changes when it gets above 8.
Combining Mr. Koga's and my experiences, the method is something like:
A final note; anything with protein in it will decompose and produce ammonia. I have heard of people using a pinch a day of fish food and even milk to provide an ammonia source for cycling the tank. You might consider these options. Also, if you have a cycled tank that you want to keep going through a period with no fish in it, try just adding a bit of fish food daily.