the best; as, if we are to use gas economically, the flame must be the same shape as the vessel to be treated. We may now consider the heating of blanks for stamping, hardening the points of spindles, finishing the ends of umbrella tips, and work where a small article, or a small part of any article, has to be heated to a high temperature with speed and certainty. For these a long and narrow flame is necessary, and I may mention that in cases where a high speed of delivery is required, and a small part only has to be heated, such as, for instance, in the hardening of the points of spindles for cotton machinery, I have made burners giving a flame of exceedingly high temperature only ¼ inch wide and five feet long. This flame is produced by the assistance of a blast of air, and is of sufficiently high temperature to fuse the spindle in a few minutes. The points only project over the flame, and the spindles are carried mechanically at such a speed that at the end of the five feet traverse they are red hot, and drop into water. More than one hundred are in the flame at once, lying side by side. For heating blanks for stamping, the furnace bar-burner is perfectly suited, and in this work the chute supplying the blanks to the machine should be made of two fireclay sides, with an opening for the flame between the chute and flame being placed at a sharp angle, to prevent risk of the blanks sticking or overriding each other. A blowpipe may also be used with good effect, as shown in the above engraving, and in many cases it is preferable and much easier to manage. In some cases the direct contact of the flame would spoil the articles to be heated, and instead of the arrangement mentioned, a tube of iron, fireclay, or other suitable material is heated, and the articles are passed through it. This system of continuous feed, through a tube, has been applied to the firing of small articles of pottery, and might possibly be well adapted, among other things, to the production of gas-burners. [Illustration: FIG. 4.] Where the contact of air with the heated articles is injurious, many plans have been tried to keep the ends closed as much as possible, but I believe no more perfect and simple seal against the admission of air can be devised than to turn a jet of pure gas, unmixed with air, into each end of the tube. This is an absolute seal against the entry of oxygen in an uncombined state; free oxygen cannot exist at a very high temperature in the presence of coal gas.