The talk turned to precautions taken against accidents. In all powder-mills the workmen are required to change their clothes before entering the buildings, and to put on rubber-soled shoes. There must be no bit of metal about a man's person, no iron nail or buckle, nothing that could strike fire; and of course the workman who would bring a match on the premises would be counted worse than an assassin.

"Just the same, though, matches get into the works once in a while," remarked the young packer. "I found a piece of a match one day in a tub of dynamite; it had the head on, too. Say, it's bad enough to find buttons and pebbles, but when I saw that match-head—well, it made me weak in the knees."

This brought back the old question, When does dynamite explode, and when does it not explode? I mentioned the red-spark theory.

"I think that's correct," agreed the packer. "I've watched 'em burn old dynamite-boxes, and if there are iron nails in the boxes they explode as soon as the nails get red-hot; if there are no nails, they don't explode."

"You mean empty boxes?" I asked.

"Certainly; but there's nitroglycerin in the wood, lots of it. It oozes out of the dynamite, especially on a hot day, and soaks into everything. Why, I suppose there's enough nitroglycerin in the overalls I wear to blow a man into—well, I wouldn't want to lay 'em on an anvil and give 'em a whack with a sledge."

There was a certain novelty to me in the thought of a pair of old overalls exploding; but I was soon to hear of stranger things. By this time other workmen had drawn up chairs, and were ready now with modest contributions from their own experience.

"Tell ye a queer thing," said one man. "In that explosion the other day,—I mean the freezing-house,—a car loaded with powder [dynamite] had just passed, not a minute before the explosion. Lucky for the three men with the car, wasn't it? But what gets me is how the blast, when it came, blew the harness off the horse. Yes, sir; that's what it did—clean off; and away he went galloping after the men as hard as he could leg it. Nobody touched a buckle or a strap. It was dynamite unhitched that animal."

"Dynamite did another trick that day," put in a tall man. "It caught a bird on the wing. Dunno whether 'twas a robin or a swaller, but 'twas a bird, all right. Caught it in a sheet of tin blowed off the roof, an' jest twisted that little bird all up as it sailed along, and when it struck the ground, there was the bird fast in a cage made in the air out of a tin roof. Alive? Yes, sir, alive; and that shows how fast dynamite does business."

So the talk ran on, with many little details of explosions. The expert explained that the air waves of a great concussion move along with crests and troughs like water waves, and the shattering effect comes only at the crests, so that all the windows might be broken in a house, say, half a mile from an explosion, and no windows be broken in a house two hundred yards nearer. The first house would have been smitten by a destructive wave crest, the second passed over by a harmless wave trough. And, by the way, when windows are broken by these blasts of concussion, it appears that they are usually broken outward, not inward, and that the fragments are found on the ground outside the house, not on the floors inside. The reason of this is that the concussion waves leave behind them a partial vacuum, and windows are broken by the air inside houses rushing out.