little
moisture, and a recognisable quantity of nitric peroxide. These may be
removed by heating the tube and drawing filtered air through it, but not
by washing, as it is difficult to obtain water which leaves no residue
on the silica.

=Making larger tubes and other apparatus of Silica.=--In order to
convert a small bulb of silica into a larger one or into a large tube,
proceed as follows:--Heat one end of a fine rod of silica and apply it
to the bulb so as to form a ring as shown in the figure. Then heat the
ring and the end of the bulb till it softens, and expand the end by
blowing. If this process is repeated, the bulb first becomes ovate and
then forms a short tube which can be lengthened at will, but the most
convenient way to obtain a very long tube is to make several shorter
tubes of the required diameter, and say 200 to 250 mm. in length, and to
join these end to end. It does not answer to add lumps of silica to the
end of the bulb, for the sides of the tube made in this way become too
thin, and blow-holes are constantly formed during the making of them.
These can be mended, it is true, but they spoil the appearance of the
work.

[Illustration]

Tubes made in the manner described above are thickened by adding rings
of silica and blowing them when hot to spread the silica. If a
combination of several jets is employed, very large tubes can be
constructed in this way. One of Messrs. Baird and Tatlock's workmen
lately blew a bulb about 5 cm. in diameter, and it was clear that he
could have converted it into a long cylindrical tube of equal diameter
had it been necessary to do so.

Very thin tubes of 1.5 cm. diameter, and tubes of considerable thickness
and of equal size, are easily made after some practice, and fine
capilliaries and millimetre tube can be made with about equal readiness.

If a very fine tube of even bore is required, it may be drawn from a
small thick cylinder after a little practice.

When a tube becomes so large that it cannot be heated uniformly on all
sides by rotating it in the flame, it is convenient to place a sheet of
silica in front of the flame a little beyond the object to be heated, in
order that the former may throw back the flame on those parts of the
tube which are most remote from the jet. A suitable plate may be made by
sticking together small lumps of silica rendered plastic by heat.

The silica tubes thus made can be cut and broken like glass, they can be
joined toge