of the bulb, and blow out the
desired hole where the tube was drawn off. The whole bulb must generally
be reheated and blown a little at the end of the process, and well
annealed.

[Illustration: FIG. 13.--_a_, Kjeldahl trap; _b_, suction pump on
smaller tubing.]

The suction pump can also be made on 1/4-inch tubing, and one joint
saved if desired, by constricting the tube to form the raceway for the
water and air, as shown in _b_, Fig. 13. (See page 10 for method.) But
it is more difficult to make a square joint on such small tubing.

CHAPTER V

MODIFIED METHODS AND SPECIAL OPERATIONS

CAPILLARY TUBING

This is commonly used in many forms of apparatus for gas analysis, and
one is often called upon to join two pieces or to make a tee on it. The
methods are nearly the same as with other tubing, except that more care
and patience are required. The work must be done much more slowly on
account of the thickness of the walls, and open ends of the tube must
always be enlarged before joining them to anything. This is best done by
carefully sealing the end and then blowing, with several suitable
reheatings, to form a pear-shaped bulb as in _a_, Fig. 14. The end of
this is then heated and blown off, and the piece is ready to be joined
to another similar end, or to a piece of ordinary tubing if desired. The
joints are best not blown too much, as thick walls shrink very slowly.
Much may be done by gently pushing the tube together or pulling it apart
in the flame, to remove lumps and irregularities. It is necessary that
the bore of the joint be approximately that of the main tube, and care
must be taken that the latter is not constricted at the point where the
joint begins.

[Illustration: FIG. 14.--Capillary tubing.]

Especial care must be taken to warm the tube slowly when starting and
cool it slowly when through, as the thick walls frequently crack if not
carefully handled. For this reason the whole neighborhood of the joint
must be heated somewhat so that there may not be stresses set up between
the heated and unheated portions.

In making the tee (_b_, Fig. 14) the inability to blow the joint makes
itself decidedly felt, but if the side tube is properly enlarged as
previously described, a good joint can be made by alternately pulling
and pushing on the end of the side tube, and shrinking well.

Very fine capillary tubing should be blown with a rubber bulb instead of
the mouth, so as not to get moisture in