be turned up nicely and the necks ground
out by a copper or iron cone and emery. The stoppers are rotated on a
lathe at quite a slow speed, say 30 or 40 feet per minute, and the
necks are held against them, as described in the section dealing with
this art. The stoppers must in this case be finished with "two
seconds" emery, and lastly with pumice dust and water (see chapter on
glass-grinding).

Unless the stoppers fit exceedingly well trouble will arise from the
mercury (which is poured into the thistle heads to form a seal) being
forced downwards into the pump by atmospheric pressure.

The joints between the three fall tubes and the single exhaust main
are easily made when the tubes are finally mounted, the hooked nozzle
of the oxygas blow-pipe being expressly made for such work.

It is, on the whole, advisable to make the pump of flint glass, or at
all events the air-trap tube and the fall tubes. A brush flame from
the larger gas tube of the single blowpipe table is most suitable for
the work of bending the tubes. The jointing of the long, narrow bore
fall tubes is best accomplished by the oxygas flame, for in this way
the minimum of irregularity is produced; the blowing tubes will of
course be required for the job, and the narrow tubes must be well
cleaned to begin with.

The air trap is an important though simple part of the pump. Its
shoulder or fall should stand rather higher than the shoulders of the
fall tubes, so that the mercury may run in a thin stream through a
good Torricellian vacuum before it passes down to the fall tubes.
This is easily attained by regulating the main mercury supply at the
pinch cock situated between the tube from the upper reservoir and the
air-trap tube, the other cocks being almost wide open.

It might be thought that the mercury would tend to pick up air in
passing through the rubber connections to the fall tubes, but I have
not found this to be the case in practice. There is, of course, no
difficulty in eliminating the rubber connections between the fall
tubes and the mercury supply from the air trap, but it impresses a
greater rigidity on the structure and, as I say, is not in general
necessary. It must not be forgotten that the mercury always exercises
considerable pressure on the rubber joints, and so there is little
tendency for gas to come out of the rubber.

The thistle funnels at the head of the fall tubes provide a simple and
excellent means of cleaning the f