make Hofman's Apparatus for the electrolysis of water_ (Fig. 27).

Take two tubes about 35 cm. in length, and 14 mm. in diameter for _AA_,
join taps _TT_ to the end _B_ of each of them, draw out the other end,
as shown at _D_, after sheets of platinum foil with wires attached to
them[12] have been introduced into the tubes, and moved by shaking to
_BB_. Then allow the platinum wires to pass through the opening _D_ left
for the purpose, and seal the glass at _D_ round the platinum as at _E_.
Pierce the tubes at _JJ_, and join them by a short piece of tube _K_,
about 14 mm. in diameter, to which the tube _T_, carrying the reservoir
_R_, has been previously united. _R_ may be made by blowing a bulb from
a larger piece of tube attached to the end of _T_. The mouth _M_ of the
reservoir being formed from the other end of the wide tube afterwards.
One of the taps can be used for blowing through at the later stages.
Each joint, especially those at _JJ_, must be annealed after it is
blown. Some operators might prefer to join _AA_ by the tube _K_ in the
first instance, then to introduce the electrodes at _E_ and _D_. In some
respects this plan would be rather easier than the other, but, on the
whole, it is better to make the joints at _JJ_ last in order, as they
are more apt to be broken than the others during the subsequent
manipulations.

[12] Red-hot platinum welds very well. The wire may be joined to the
sheet of foil by placing the latter on a small piece of fire-brick,
holding the wire in contact with it at the place where they are to be
united, directing a blow-pipe flame upon them till they are at an
intense heat, and smartly striking the wire with a hammer. The blow
should be several times repeated after re-heating the metal.

2. I have before me the vacuum tube shown by Fig. 28, in which the
dotted lines relate to details of manipulation only.

[Illustration: FIG. 28.]

It is usually possible to detect the parts of which a piece of apparatus
has been built up, for even the best-made joints exhibit evidence of
their existence. Thus, although I did not make the tube that is before
me, and cannot therefore pretend to say precisely in what order its
parts were made and put together, the evidence which it exhibits of
joints at the dotted lines _A_, _B_, _C_, _D_, _E_, _F_, enables me to
give a general idea of the processes employed in its construction, and
to explain how a similar tube might be constructed. I should advi