The primers in the ammunition manufactured at the Frankford Arsenal had given ordinarily satisfactory results in 12 years of peace-time use. The flame charge in this primer contained sulphur, potassium chlorate, and antimony sulphide. Produced under normal conditions, with plenty of time for drying, this primer was satisfactory. But sulphur when oxidized changes to an acid extremely corrosive to metal parts, and oxidized primers are liable not to function perfectly. Heat and moisture accelerate the change of sulphur to acid; and if there happens to be bromate in the potassium chlorate of the priming charge, the change is even more rapid. An investigation of the Frankford Arsenal showed that these very elements were present. Because of the haste of production of cartridges, too much moisture had been allowed to get into the arsenal dry houses. The potassium chlorate was also found to contain appreciable quantities of bromate. The condition was remedied by adopting another primer composition. And then, to play doubly safe, the Government specifications were amended to prevent the use of potassium chlorate containing more than 0.01 per cent of bromate. However, this condemned ammunition was but a trifling fraction of the total output or even of the production then going on. The primers used by the various private manufacturers of ammunition functioned satisfactorily. While we were not rigid in our specifications for the bulk of the service ammunition, in one respect we were most meticulous, and this was in respect to the ammunition used by the machine guns mounted on our airplanes. For these weapons we created an A-1 class of service .30-caliber cartridges, since it was highly important that there be no malfunctioning of ammunition in the air. Every cartridge of this class had to be specially gauged throughout its manufacture. This care resulted in a slower production of airplane cartridges than that of those for use on the ground, but we always had enough for our needs. Until we went to war with Germany our Army had known only the cartridge firing the hard-jacketed lead bullet. But we entered a conflict in which several novel sorts of small-arms projectiles were in familiar use; and it became necessary for us to take up the manufacture of these strange missiles at once. These included such special types as tracer bullets to indicate the path of fire in the air, incendiary bullets for setting on fire observation balloons, hostile planes, and dirigible airships, and, finally, armor-piercing bullets for use against armor plate with which airplanes and tanks are equipped. We had developed none of these in this country before the war, except that in the Frankford Arsenal our designers had done some little experimental work with armor-piercing ammunition, in fact carrying it to the point of an efficient design.