of an oval-shaped India rubber ball,
fitted with a mouthpiece. The diameter is about 21/4 inches by 3 inches, its
weight is two ounces, and it is so small that it can be carried without any
inconvenience in the coat or even in the waistcoat pocket. Its capacity is
such that all the air within it may be expelled by the compression of one
hand.

The mouthpiece is made to fit a tube in the bottom of the lamp, and when
pressed against the India rubber ring on the ball-flange, a perfectly tight
joint is made, which prevents the admission of any external air. The tube
in the bottom of the lamp is carried within a short distance of the height
of the wick-holder. It is covered at the upper end with gauze, besides
being fitted with other thicknesses of gauze at certain distances within
the tube; and if it be found desirable to further protect the flame against
strong currents of air, a small valve may be placed at the inlet, as shown
in the drawing. This valve is made of sufficient weight to resist the force
of a strong current, and is only lifted from its seat by the pressure of
the hand on the mouthpiece. It will be apparent from the small size and
elasticity of the detecter that the test can easily be made with one hand,
and when the ball is allowed to expand a vacuum is formed within it, and a
sample of the atmosphere drawn from the breaks, cavities, or highest parts
of the roof, or, of course, any portion of the mine. When the sample is
forced through the tube near the flame, gas if present at once reveals
itself by the elongation of the flame in the usual way, at the same time
giving an additional proof by burning with a blue flame on the top of the
test tube. If gas is not present, the distinction is easily seen by the
flame keeping the same size, but burning with somewhat greater brightness,
owing to the increased quality of oxygen forced upon it.

I venture to claim for this method of detecting fire-damp among other
advantages: 1. The detecter, on account of its size, can be placed in a
break in the roof where an ordinary lamp--even a small Davy--could not be
put, and a purer sample of the suspected atmosphere is obtained than would
be the case even a few inches below the level of the roof, 2. The obtaining
and testing of a sample in the manner above described takes away the
possibility of an explosion, which might be the result if a lamp with a
defective gauze were placed in an explosive atmosphere. No one knows how
ma