ock bottle every time a bubble is
to be blown, but have a small working bottle. Never put any back into
the stock. In making the liquid _do not warm or filter it_. Either will
spoil it. Never leave the stoppers out of the bottles or allow the
liquid to be exposed to the air more than is necessary. This liquid is
still perfectly good after two years' keeping. I have given these
directions very fully, not because I feel sure that all the details are
essential, but because it exactly describes the way I happen to make it,
and because I have never found any other solution so good. Castille
soap, Price's glycerine, and rain-water will almost certainly answer
every purpose, and the same proportions will probably be found to work
well.

_Rings for Bubbles._

These may be made of any kind of wire. I have used tinned iron about
one-twentieth of an inch in diameter. The joint should be smoothly
soldered without lumps. If soldering is a difficulty, then use the
thinnest wire that is stiff enough to support the bubbles steadily, and
make the joint by twisting the end of the wire round two or three times.
Rings two inches in diameter are convenient. I have seen that dipping
the rings in melted paraffin is recommended, but I have not found any
advantage from this. The nicest material for the light rings is thin
aluminium wire, about as thick as a fine pin (No. 26 to 30, B. W. G.),
and as this cannot be soldered, the ends must be twisted. If this is not
to be had, very fine wire, nearly as fine as a hair (No. 36, B. W. G.),
of copper or of any other metal, will answer. The rings should be wetted
with the soap mixture before a bubble is placed upon them, and must
always be well washed and dried when done with.

_Threads in Ring._

There is no difficulty in showing these experiments. The ring with the
thread may be dipped in the soap solution, or stroked across with the
edge of a piece of paper or india-rubber sheet that has been dipped in
the liquid, so as to form a film on both sides of the thread. A needle
that has also been wetted with the soap may be used to show that the
threads are loose. The same needle held for a moment in a candle-flame
supplies a convenient means of breaking the film.

_Blow out Candle with Soap-Bubble._

For this, the bubble should be blown on the end of a short wide pipe,
spread out at one end to give a better hold for the bubble. The tin
funnel supplied with an ordinary gazogene answers perfec