Sec. 8. As against the advantages enumerated, there are some drawbacks. The one which will first strike the student is the tendency of the glass to become reduced in the flame of the blow-pipe. This can be got over by proper adjustment of the flame, as will be explained later on. A more serious drawback in exact work is the following. In making a joint with lead glass it is quite possible to neglect to fuse the glass completely together at every point; in fact, the joint will stand perfectly well even if it be left with a hole at one side, a thing which is quite impossible with soft soda glass, or is at least exceedingly unusual. An accident of this kind is particularly likely to happen if the glass be at all reduced. Hence, if a joint does not crack when cold, the presumption is, in the case of soda glass, that the joint is perfectly made, and will not allow of any leak; but this is not the case with flint glass, for which reason all joints between flint glass tubes require the most minute examination before they are passed. If there are any air bubbles in the glass, especial care must be exercised. Sec. 9. Hard or Bohemian, Glass. This is, of course, used where high temperatures are to be employed, and also in certain cases where its comparative insolubility in water is of importance. It is very unusual for the investigator to have to make complicated apparatus from this glass. Fused joints may be made between hard glass and flint glass without using enamel, and though they often break in the course of time, still there is no reason against their employment, provided the work be done properly, and they are not required to last too long. Sec. 10. On the Choice of Sizes of Glass Tube. It will be found that for general purposes tubes about one-quarter inch in inside diameter, and from one-twentieth to one-fortieth of an inch thick, are most in demand. Some very thin soda glass of these dimensions (so-called "cylinder" tubes) will be found very handy for many purposes. For physico-chemical work a good supply of tubing, from one-half to three-quarters of an inch inside diameter, and from one-twentieth to one-eighth inch thick, is very necessary. A few tubes up to three inches diameter, and of various thicknesses, will also be required for special purposes. Thermometer and "barometer" tubing is occasionally required, the latter, by the way, making particularly bad barometers. The thermometer tubing should be of all sizes of bore, from the finest obtainable up to that which has a bore of about one-sixteenth of an inch. Glass rods varying from about one-twentieth of an inch in diameter up to, say, half an inch will be required, also two or three sticks of white enamel glass for making joints.