to leave very finely-serrated edges, will be found useful for cutting
glass tubes. Such a file should be used almost as a knife is used for
cutting a pencil in halves.

The simple methods just described are too violent to be applied to
delicate apparatus, too tedious when employed upon the largest tubes,
and very difficult to apply when the tube to be cut is very thin, or too
short to permit the operator to get a good grip of it on either side of
the file mark. In such cases, one or other of the following methods will
be useful:--

1. Make a scratch with a file, and touch it with the end of a _very
small_ piece of glass drawn out and heated at the tip to its melting
point. It is important that the heated point of glass be very small,
or the fracture is likely to be uneven, or to spread in several
directions. Also, it is best to use hot soda glass for starting cracks
in tubes of soda glass, and lead glass for doing so in lead glass
tubes. If the crack does not pass quite round the tube, you may pull
it asunder, as previously described, or you may bring the heated piece
of glass with which the crack was started to one end of the crack, and
slowly move it (nearly touching the glass) in the required direction;
the crack will extend, following the movements of the hot glass.
Instead of hot glass, pastils of charcoal are sometimes employed for
this purpose. They continue to burn when once lighted, and there is
no need to re-heat them from time to time. They should be brought as
close to the glass as is possible without touching it, and, when no
longer needed, should be extinguished by placing the lighted end under
sand, or some other incombustible powder, for they must not be wetted.

2. A method much practised by the makers of sheet glass, and suitable
for large objects, is to wrap a thread of hot glass round the tube, at
once removing it, and touching any point of the glass which the thread
covered with water or a cold iron, when a crack will be started and will
pass round the glass where it was heated by the thread.

3. Tubes which are large and slightly conical may have a ring of red-hot
iron passed over them till it comes into contact with the glass, then,
the iron being removed, and a point on the heated glass being at once
touched with cold iron as before, it will break as desired. Or a string,
moistened with turpentine, may be loosely twisted round the tube, and
the turpentine ignited, afterwards the application