hin less than two inches of the point where the second seal is to be
made. If the substance is a liquid it can more conveniently be
introduced at a later stage.

Now bring the tube into the blowpipe flame gradually, and rotate it,
while heating, at the place where it is to be closed. Allow the glass to
soften and commence to run together until the diameter of the tube is
reduced to about half its original size. Remove from the flame and draw
the ends apart, this should give a long, thick extension as shown by
_f_, Fig. 4. If any liquid is to be introduced, it may now be done by
inserting a thin rubber or other tube through the opening and running
the liquid in. A glass tube should be used with caution for introducing
the liquid, as any hard substance will tend to scratch the inside of the
glass and cause cracking. The final closure is made by melting the
drawn-out extension in the blowpipe flame; the finished seal being shown
by _g_, Fig. 4.

If the sealed tube has to stand internal pressure, it is desirable to
allow the glass to thicken somewhat more before drawing out, and the
bottom seal should also be made thicker. For such a tube, and especially
when it has to stand heating, as in a Carius determination of chlorine,
each seal should be cooled very slowly by rotating it in a gas flame
until the surface is covered with a thick layer of soot, and it should
then be placed aside in a position where the hot glass will not come in
contact with anything, and where it will be screened from all draughts.

_Joining Tube._--We will now consider the various forms of join in glass
tubing which are met with in the laboratory. First, as being easiest, we
will deal with the end-to-end joining of two tubes of similar glass.
_a_, _b_, and _c_, Fig. 5, illustrate this. One end of one of the tubes
should be closed, a lip should be turned out on each of the ends to be
joined, and both lips heated simultaneously until the glass is
thoroughly soft. Now bring the lips together gently, until they are in
contact at all points and there are no places at which air can escape;
remove from the flame, and blow slowly and very cautiously until the
joint is expanded as shown in _b_, Fig. 5. Reheat in the flame until
the glass has run down to rather less than the original diameter of the
tube, and give a final shaping by re-blowing. The chief factors of
success in making such a join are, thorough heating of the glass before
bringing the two tubes