his direct-line
lamp arrangement is not a good one where the lines vary widely in length
and resistance. An incandescent lamp, as is well known, must not be
subjected to too great a variation in current. If the current that is
just right in amount to bring it to its intended degree of illumination
is increased by a comparatively small amount, the life of the lamp will
be greatly shortened, and too great an increase will result in the
lamp's burning out immediately. On the other hand, a current that is too
small will not result in the proper illumination of the lamp, and a
current of one-half the proper normal value will just suffice to bring
the lamp to a dull red glow. With lines that are not approximately
uniform in length and resistance the shorter lines would afford too
great a flow of current to the lamps and the longer lines too little,
and there is always the danger present, unless means are taken to
prevent it, that if a line becomes short-circuited or grounded near the
central office, the lamp will be subjected to practically the full
battery potential and, therefore, to such a current as will burn it out.
One of the very ingenious and, we believe, promising methods that has
been proposed to overcome this difficulty is that of the iron-wire
ballast, alluded to in Chapter III. This, it will be remembered,
consists of an iron-wire resistance enclosed in a vacuum chamber and so
proportioned with respect to the flow of current that it will be
subjected to a considerable heating effect by the amount of current that
is proper to illuminate the lamp. As has already been pointed out,
carbon has a negative temperature coefficient, that is, its resistance
decreases when heated. Iron, on the other hand, has a positive
temperature coefficient, its resistance increasing when heated. When
such an iron-wire ballast is put in series with the incandescent lamp
forming the line signal, as shown in Fig. 308, it is seen that the
resistance of the carbon in the lamp filament and of the iron in the
ballast will act in opposite ways when the current increases or
decreases. An increase of current will tend to heat up the iron wire of
the ballast and, therefore, increase its resistance, and the ballast is
so proportioned that it will hold the current that may flow through the
lamp within the proper maximum and minimum limits, regardless of the
resistance of the line in which the lamp is used. This arrangement has
not gone into wide use