When no more liquid can be caused to drain away, the tube may be dried by heating it along its length, beginning at the top (to get the advantage of the reduction of surface tension), and so on all down. It will then be possible to mop up a little more of the rinsing liquid. When the tube is nearly dry a loose plug of cotton wool may be inserted at the bottom. The wool must be put in so that the fibres lie on an even surface inside the tube, and the wool must be blown free from dust. Ordinary cotton wool is useless, from being dusty and the fibres short, and the same remark applies to wadding. Use nothing but what is known as "medicated" cotton wool with a good long fibre. The tube will usually soon dry of itself when the cover is lifted an inch or so. If water has been used, the air-current may be assisted by means of the water-pump, the air being sucked from the top, so that the wool has an opportunity of acting as a dust filter; a very slow stream of air only must be employed. For connecting the tube to the pump, a bit of India-rubber tube about an inch in diameter, with a bore of about one-eighth of an inch, may be employed. The end of the rubber tube is merely pressed against the edge of the glass. These remarks apply, with suitable modification, to all kinds of finished apparatus having two openings. For flasks and so on, it is convenient to employ a blowing apparatus, dust being avoided by inserting a permanent plug of cotton wool in one of the leading tubes. The efficiency of this method is greatly increased by using about one foot of thin copper tube, bent into a helix, and heated by means of a Bunsen burner; the hot air (previously filtered) is passed directly into the flask, bottle, or whatever the apparatus may be. This has proved so convenient that a copper coil is now permanently fastened to the wall in one of the rooms of my laboratory. The above instructions indicate greater refinement than is in general necessary or proper for tubes that have to be afterwards worked by the blow-pipe. In the majority of cases all that is necessary is to remove the dust, and this is preferably done by a wad of cotton waste (which does not leave shreds like cotton wool), followed by a bit of bibulous filter paper. I would especially warn a beginner against neglecting this precaution, for in the process of blowing, the dust undergoes some change at the heated parts of the apparatus, and forms a particularly obstinate kind of dirt. In special cases the methods I have advocated for removing dirt and drying without covering the damp surfaces with dust are inadequate, but an experimenter who has got to that stage will have nothing to learn from such a work as this. Sec. 14. The Blow-pipe.