The separate air and gas taps must be left for permanent regulation, and must not be used to turn the supply on or cut it off. In some respects this blow-pipe will be found more easy to manage than an oxygas blow-pipe, for the glass is not so readily brought to the very fluid state, and this will often enable a beginner who proceeds cautiously to do more than he could with the more powerful instrument. Though I have mentioned glass nozzles for the air supply, there is no difficulty in making nozzles of brass. For this purpose let the end of a brass tube of about one-eighth of an inch diameter be closed by a bit of brass wire previously turned to a section as shown (Fig. 6), and then bored by a drill of the required diameter, say -.035 inch. It is most convenient to use too small a drill, and to gradually open the hole by means of that beautiful tool, the watchmaker's "broach." The edges of the jet should be freed from burr by means of a watchmaker's chamfering tool (see Saunier's Watchmaker's Hand-book, Tripplin, 1882, p. 232, Sec. 342), or by the alternate use of a slip of Kansas stone and the broach. Fig. 6. The construction of this blow-pipe is so simple, that in case any one wishes to use a brush flame, he can easily produce one simply by changing his air jets to bits of the same size (say one-eighth to one-sixteenth of an inch) tubing, cut off clean. To insure success, the ends of the tubes must be absolutely plane and regular; the slightest inequality makes all the difference in the action of the instrument. If a jet is found to be defective, cut it down a little and try again; a clean-cut end is better than one which has been ground flat on a stone. The end of a tube may, however, be turned in a manner hereafter to be described so as to make an efficient jet. Several trials by cutting will probably have to be made before success is attained. For this kind of jet the air-pressure must be greatly increased, and a large Fletcher's foot-blower or, better still, a small double-action bellows worked with vigour will be found very suitable. A fitting for this auxiliary blow-pipe is shown in Fig. 5 at B. Professor Roentgen's discovery has recently made it necessary to give more particular attention to the working of soft soda glass, and I have been obliged to supplement the arrangements described by a table especially intended for work with glass of this character. The arrangement has proved so convenient for general work that I give the following particulars. The table measures 5 feet long, 2 feet 11 inches wide, and is 2 feet 9 inches high. Fig. 7.