burn very well, you will often see the man poking it with a
pin. The carbon given off from the naphtha is very disposed to choke up
the little hole through which the naphtha runs into the cup, and the
costermonger pushes a pin into the little hole to allow the free passage
of the naphtha. That, then, is the mechanism of this beautiful lamp of
the Whitechapel traders, known as Halliday's lamp.

Now I go to another point: having obtained the gas, I must set fire to
it. It is important to note that the temperature required to set fire
to different gases varies with the gas. For instance, I will set free in
this bottle a small quantity of gas, which fires at a very low
temperature. It is the vapour of carbon disulphide. See, I merely place
a hot rod into the bottle, and the gas fires at once. If I put a hot rod
into this bottle of coal gas, no such effect results, since coal gas
requires a very much higher temperature to ignite it than bisulphide of
carbon gas. I want almost--not quite--actual flame to fire coal gas. But
here is another gas, about which I may have to say something directly,
called marsh gas (the gas of coal-mines). This requires a much higher
temperature than even coal gas to fire it. I want you to understand that
although all gases require heat to fire them, different gases ignite at
very different temperatures. Bisulphide of carbon gas, _e. g._, ignites
at a very low temperature, whilst marsh gas requires a very high
temperature indeed for its ignition. You will see directly that this is
a very important fact. Sulphur gas ignites fortunately at a fairly low
temperature, and that is why sulphur is so useful an addition to the
wood splint by which to get fire out of the tinder-box.

[Illustration: Fig. 31.]

And here I wish to make a slight digression in my story. I will show you
an experiment preparatory to bringing before you the fact I am anxious
now to make clear. I have before me a tube, one half of which is brass
and the other half wood. I have covered the tube, as you see, with a
tightly-fitting piece of white paper. The whole tube, wood and brass,
has been treated in exactly the same manner. Now I will set fire to some
spirit in the trough I have here, and expose the entire tube to the
action of the flame. Notice this very curious result, viz. that the
paper covering the brass portion of the tube does not catch fire,
whereas the paper covering the wood is rapidly consumed (Fig. 31). You
see the