f the glass globe into a
metallic cup, and to provide it with a tube or stop-cock for exhaustion
by means of a hand-pump. Lodyguine, Konn, Kosloff, and Khotinsky,
between 1872 and 1877, proposed various ingenious devices for perfecting
the joint between the metal base and the glass globe, and also provided
their lamps with several short carbon pencils, which were automatically
brought into circuit successively as the pencils were consumed. In 1876
or 1877, Bouliguine proposed the employment of a long carbon pencil, a
short section only of which was in circuit at any one time and formed
the burner, the lamp being provided with a mechanism for automatically
pushing other sections of the pencil into position between the contacts
to renew the burner. Sawyer and Man proposed, in 1878, to make
the bottom plate of glass instead of metal, and provided ingenious
arrangements for charging the lamp chamber with an atmosphere of pure
nitrogen gas which does not support combustion.

These lamps and many others of similar character, ingenious as they
were, failed to become of any commercial value, due, among other things,
to the brief life of the carbon burner. Even under the best conditions
it was found that the carbon members were subject to a rapid
disintegration or evaporation, which experimenters assumed was due to
the disrupting action of the electric current; and hence the conclusion
that carbon contained in itself the elements of its own destruction, and
was not a suitable material for the burner of an incandescent lamp. On
the other hand, platinum, although found to be the best of all materials
for the purpose, aside from its great expense, and not combining with
oxygen at high temperatures as does carbon, required to be brought
so near the melting-point in order to give light, that a very slight
increase in the temperature resulted in its destruction. It was assumed
that the difficulty lay in the material of the burner itself, and not in
its environment.

It was not realized up to such a comparatively recent date as 1879 that
the solution of the great problem of subdivision of the electric current
would not, however, be found merely in the production of a durable
incandescent electric lamp--even if any of the lamps above referred to
had fulfilled that requirement. The other principal features necessary
to subdivide the electric current successfully were: the burning of an
indefinite number of lights on the same circuit; ea