One of the boys accidentally discovered that when laid upon a stone and the edge of the hat hit with a hammer, it would crack, so they took it to a blacksmith's shop, where they could have some fun by hammering it on an anvil. At the first blow the old hat exploded. Two of the boys were killed outright, and two more were badly injured. The blacksmith at the time of the accident, happened to be standing outdoors, which thereafter constituted his blacksmith shop until he could rebuild. A DROP TOO MUCH Professor Mowbray, who made the nitroglycerin for the Hoosac Tunnel and afterward served the American Xylonite Company many years as consulting chemist, conceived the idea that he could make a very powerful smokeless gunpowder by the use of nitroglycerin merely absorbed by fibrous guncotton and rolled into pellets. He had at the time a young assistant chemist at work for him, who has now become a man of much wealth and prominence in New York. The assistant prepared some of the pellets under Mowbray's directions, loaded them into a rifle under wad and ball, and fired at a target made of several layers of pine boards. But the pellets did not seem to give the bullet the required penetration. Mowbray suggested remedying this defect by adding a little more nitroglycerin, which was done. The young chemist demurred a little. Still, he did as instructed--loaded and fired the piece again, with but little better results. This time, however, the breech mechanism stuck, and was opened with difficulty. Mowbray said that there was but one thing to do, and that was to add a few more drops of nitroglycerin. It occurred to the young chemist that this sort of gunpowder came pretty near being dynamite, and he declined to fire the piece the next time, and was deaf to all entreaties of the Professor. As a compromise, the gun was rigged up on a rest, pointing at the target; a string was attached to the trigger, which the assistant, standing behind a barricade, pulled. This time, there was considerable penetration of the target, and the walls of the building where the test took place were penetrated in many places, not with the bullet, but with the fragments of the exploded weapon. Mowbray, hearing the report, ran out and ventured the suggestion that he guessed he must have got in a drop too much of nitroglycerin. A CLOSE CALL I had one very close call while conducting a sand test of the Maxim gun at Annapolis, where the Naval Proving Grounds were formerly located. The gun had passed through all of the regular tests satisfactorily, and it was then suggested to try if sand enough could be put into the mechanism box to block it and prevent its firing.