His first object was to reduce iron oxides. First, he knew that aluminum has a great affinity for oxygen, or, in other words, when the two are heated will absorb oxygen like a sponge will absorb water, only more forcibly and more violently than any such comparison even faintly suggests. In yet other words, aluminum wants oxygen more than any other metal does. Of course no chemical changes would occur if a piece of iron oxide and a piece of aluminum were set side by side, any more than we would have gunpowder if we set a chunk of saltpetre, a chunk of sulphur, and a chunk of charcoal all in a row. The iron oxide and the aluminum would have to be mixed by cutting or filing them into small pieces and making a coarse powder. Still nothing would happen without heat to start it. If you collected some flakes of iron oxide in the palm of your hand they wouldn't look to you like very promising material for a bonfire, and you wouldn't be in any danger of an explosion, but you would have something in your hand that would burn, nevertheless. If you sprinkled your iron filings over a gas flame, Welsbach burner, or over a common lamp chimney the heat would cause them to splutter and fly out with all the brilliancy you know so well when the blacksmith gives the redhot horseshoe the first pound. Of course Doctor Goldschmidt knew all this, just as he knew that the way the aluminum would take the oxygen away from the iron oxide was through heating the coarse powder of filings to a very high temperature. But this was attended with serious troubles and many times the German scientist came near losing his life in explosions in his laboratory. At first he failed to get the mixture hot enough and nothing happened. Bit by bit he increased the heat under the crucible containing the filings until it reached about 3,000 degrees Fahrenheit. At this point the metals were hot enough to fuse or run together and the whole thing reacted with such violence that it amounted to an explosion. What really happened was that the mass reached the temperature where the aluminum could take the oxygen from the iron oxide, and it did so with such force that an explosion resulted. Doctor Goldschmidt then saw his problem. It was that of devising some way of heating the mixture to a temperature sufficient to gain the reaction, but without an explosion. After trying everything that he could think of, he conceived the plan of leaving the crucible in the open air and starting the heat at just one point first, instead of heating the whole thing in a furnace. He did this with a pinch of ignition powder placed on the top of his pile of iron oxide and aluminum. The ignition powder was simply lighted with a match. What happened? Thermit was discovered. The heat, or reaction started at one point, gradually spread through the whole mass, and reduced it to white-hot molten material.