Iron filings, or, still better, those of steel since they contain more sulphur, give a fire that is very brilliant for the purposes of pyrotechny. As a rule, these are used as required for all varieties of fireworks. The metal to be employed should be new, since that which is rusted either fails to give sufficient brilliancy, or makes no display of sparks whatever. It should be noted that the filings possess no lasting qualities. Their value usually endures for no more than six days. The saltpeter attacks and destroys the metal, thus causing it to lose its brilliancy day by day.

The use of such filings in pyrotechny is due to the researches of Father d' Incarville, a Jesuit, who conducted a mission in Pekin. While stationed there he contrived to gather full knowledge concerning the methods by which through countless years the Chinese have produced brilliant fires, especially employed in the representation of flowers.

This preparation, carefully kept secret until the present time, is gained by reducing the iron into tiny fragments, so that the fire of the combustible composition into which they enter shall be sufficient to cause fusion. Each particle of the metal when it is fused, even though it may be no larger than a poppy seed, gives a large "flower," 12 or 15 lines in size, from a fire of utmost brilliancy, and this fire assumes various forms such as may be desired, following both the quality of the metal itself and also the shape and bulk of the grains, according as these may be round, flat, oblong, triangular, and the like. Each gives to its particular flower a different formation. This material, which Father d' Incarville names iron sand, is made out of old iron pots, or such other articles of the metal as are capable of being broken and reduced into tiny fragments on an anvil. Use may be made also of pieces too small in size to be easily broken up, for the operation may be facilitated by heating the metal, and then plunging it into a bucket of cold water. This process renders it more readily breakable. Moreover, the effect is improved when the hammer and anvil also are of iron. It is customary to spread cloths on the ground about the anvil, in order that none of the sand shall be lost. Care should be taken that no dirt be permitted to mingle with the particles of metal. When a sufficient quantity of sand has been thus secured, it is first passed through a very fine screen, in order to get rid of the useless dust. Afterward, it is passed through screens of various sizes, in order to separate it into grades as required, from those that are finest up to those as large as a radish seed.

The separate sorts are placed each by itself, to be kept in a very dry place, so that they shall not rust. It should be noted that the plunging of the hot metal into the water not only increases the facility of its reduction into the sand form, but also causes a structural or chemical alteration. There is an appreciable difference between the sand from metal thus tempered and that from other metal. The tempered sort produces flowers both larger and more brilliant. In addition, this sand will remain for a much longer time without suffering deterioration.

Since this material has no effect except when it is fused, it is evident that a greater amount of heat is required in order to make the coarser sand effective. The size of the case in making any sort of fireworks must be proportioned to meet this need of fusing the metal, and the character and amount of the loading charge also must be similarly proportioned. Such proportions must be increased very considerably to insure a due effect with the coarser forms of the iron sand, since the combustion must continue for a much longer time.

It is easily possible to observe the effect of the iron sand without troubling to prepare the firework itself. It is only necessary to throw a pinch of the sand into the flame of a candle. It becomes fused while passing through the flame, and displays its "flower forms." Ordinary filings may be tested in the same manner. Since these contain less sulphur than the sand, they merely give out sparks, similar to those thrown off by steel striking against a flint.

A firework containing this sand retains its effectiveness for about a week when the particles are small, and for about a fortnight when they are of the large size. It is to be hoped that there may yet be discovered some means of pro-tecting the iron sand from the chemical reaction of the saltpeter, which destroys it.