that we have been using for some years under the name of matches.

[Illustration: Fig. 1.--MODE OF USING THE GAS LIGHTER.]

The first of these is a portable apparatus designed for lighting gas
burners, and is based upon the calorific properties of the electric
spark produced by the induction bobbin. Its internal arrangement is such
as to permit of its being used with a pile of very limited power and
dimensions. The apparatus has the form of a rod of a length that may be
varied at will, according to the height of the burner to be lighted, and
which terminates at its lower part in an ebonite handle about 4
centimeters in width by 20 in length (Fig. 1). This handle is divided
into two parts, which are shown isolatedly in Fig. 2, and contains the
pile and bobbin. The arrangement of the pile, A, is kept secret, and all
that we can say of it is that zinc and chloride of silver are employed
as a depolarizer. It is hermetically closed, and carries at one of its
extremities a disk, B, and a brass ring, C, attached to its poles and
designed to establish a communication between the pile and bobbin when
the two parts of the apparatus are screwed together. To this end, two
elastic pieces, D and E, fit against B and C and establish a contact. It
is asserted that the pile is capable of being used 25,000 times before
it is necessary to recharge it. H is an ebonite tube that incloses and
protects the induction bobbin, K, whose induced wire communicates on the
one hand with the brass tube, L, and on the other with an insulated
central conductor, M, which terminates at a point very near the
extremity of the brass tube. The currents induced in this wire produce a
series of sparks between the tube, L, and the rod, M, which light the
gas when the extremity of the apparatus is placed in proximity with the
burner.

[Illustration: Fig. 2. MECHANISM OR THE INDUCTION SPARK GAS LIGHTER.]

The ingenious and new part of the system lies in the mode of exciting
the induced currents. When the extremity of the tube, L, is brought near
the gas burner that is to be lighted, it is only necessary to shove the
botton, F, from left to right in order to produce a _limited_ number of
sparks sufficient to effect the lighting. The motion of the button has
not for effect, as might be believed, the closing of the circuit of the
pile upon the inducting circuit of the bobbin. In fact in its normal
position, the vibrator is distant from its contact, and the closin