I guess most of us have been cleaning out tanks from newly acquired but (very) old lamps, or those given for repair. I used to go the same route as so often is described: fill the tank with paraffin, petrol or whatever solvent, ad some nuts and bolts or other small hardware, and SWIRL! Not too long ago I got a fair bit of cutting to do through concrete and sortlike products. Somehow I got to read a report about the efficiency of dry cutting versus cutting with a liquid. Most often water is used to surpress the enormous cloud of dust that otherwise is the result of cutting through concrete. But research showed water also dampened the speed of cutting: dry cutting was markedly more efficient. Last week I went to clean a 47-ish Bialaddin that looked like in a very original condition. Thanks to that enormous filler opening I could see a tapestry of rubble sitting tight on the bottom that I could not scrape with a screwdriver with much effect. I soaked that bottom with acetone, bought relatively cheap in a 5 liter can, for 2 days. I then poured out the now not-so clear acetone in a plastic bottle to settle and reuse for similar chores. With the crud in the tank still moist, and possibly softened, I now put in as many bolts, no nuts, I could fit through that big opening. The reason for this selection of bolts and no nuts, is that a bolt practically always presents a sharper edge to the surface it lays on, whereas a nut can lay perfectly flat on a surface to be scraped, doing supposedly very little in being abbrasive. I also looked for the biggest load or weight to put inside, as supposedly there would be more pressure that way on whatever edges cutting into that crud. Plus, you can shake ad infinitum with just 4 bolts, but 40 of 'em likely make for quicker results. The rattle on the metal sounded more businesslike than the mooted swish I remember from the submersed hardware approach. The end result was a volume of gray/brown granules, shaken and then rinsed out of the tank, the dried sediment of which is shown here broken up. Inspection inside the tank showed a brassy bottom with some metallic effect due to the abrasive action from the bolts. I have not measured effectiveness of shaking dry against the old wet method , but I was pleased to see this metallic surface in less than 15 minutes altogether. This Bialaddin tank is clean!