Never try to force the operation; the nail soon cools the glass, so that only a very short time is available after each heat; during this the tube should be rotated against the nail rather than the nail against the tube. Be careful not to heat a greater length of tube than is necessary, or the nail will, by its component of pressure along the tube, cause the latter to "jump up" or thicken and bulge. Both ends being prepared, and if possible, kept hot, the weld may be made as before, and the heating continued till the glass falls in to about its previous thickness, leaving a bore only slightly greater than before. It is in operations such as this that the asbestos box will be found of great use. As soon as one end of the weld is ready cool it in the flame till soot deposits, and then plunge it into the asbestos. This will cause it to cool very slowly, and renders it less likely to crack when again brought into the flame. Turned-out ends, if the glass is at all thick, are very liable to crack off on reheating, so that they must be reintroduced (into the flame) with especial care. This liability to breakage is reduced, but not eliminated, by the asbestos annealing. Figs. 18 and 19. Sec. 25. When the bore is very fine, it is best to seal off the tubes, and blow an incipient bulb near one end of each tube. These bulbs may be cooled in asbestos, and cut across when cold by means of a scratch touched at one end (Figs. 18 and 19) by a fine point of highly incandescent glass. For details of this method see p. 46, Fig. 21. Time is occasionally saved by blowing off the ends of the bulbs. The details of this process will be described when the operation of making thistle-headed tubes is dealt with. Sec. 26. When the tubes are both of large diameter, long, and very thin (cylinder tubes), a considerable amount of difficulty will be experienced. On the whole, it is best to heat each end separately till the glass thickens a little, anneal in the flame and in asbestos, and then proceed as in Sec. 22. If the ends are not quite true, it will be found that quite a thickness of glass may be "jumped" together at one side of the tubes, while the edges are still apart at the other. When this looks likely to happen, incline the tubes as if the joint were a hinge, and bend back quickly; do not simply continue to push the tubes together in a straight line, or an unmanageable lump of glass will be formed on one side.