d for aniline on the
large scale, the labours of chemists, from Unverdorben in 1826 to Hofmann
in 1843, had prepared the way for the manufacturer. It must be understood
that although Runge had discovered aniline in coal-tar, this is not the
source of our present supply, for the quantity is too small to make it
worth extracting. A mere trace of aniline is present in the tar ready
formed; from the time this base was wanted in large quantities it had to
be made by nitrating benzene, and then reducing the nitrobenzene.

The light oils of tar distillation rejected by the timber-pickling
industry now came to the front, imbued with new interest to the
technologist as a source of benzene for the manufacture of aniline. The
inauguration of this manufacture, like the introduction of steam
locomotion, is connected with a sad catastrophe. Mansfield, who first
showed manufacturers how to separate benzene and other hydrocarbons from
the light oils of coal-tar, and who devised for this purpose apparatus
similar in principle to that used on a large scale at the present time
(see Fig. 5), met with an accident which resulted in his death. In the
upper part of a house in Holborn in February 1856, this pioneer was
carrying on his experiments, when the contents of a still boiled over and
caught fire. In his endeavours to extinguish the flames he received the
injuries which terminated fatally. Applied science no less than pure
science has had its martyrs, and among these Mansfield must be ranked.

[Illustration: FIG. 5.--Mansfield's still. R the heating burner, A the
body of the still with stopcock, _i_, for running out the contents. B the
still-head kept in a cistern, C, of hot water or other liquid. The vapour
generated by the boiling of the liquid in A, partly condenses in B, from
whence the higher boiling-point portion flows back into the still. The
uncondensed vapour passes into the condensing-worm, D, which is kept cool
by a stream of water, and from thence flows into the receiver S. By
opening _m_ in the side-pipe any higher boiling-point oil condensing in
the delivery-pipe can be run back into the still.]

The operation of tar-distilling is about as unromantic a process as can be
imagined, but it must be briefly described before the subsequent
developments of the industry can be appreciated properly. It has already
been explained that the tar is a complex mixture of many different
substances. These various compounds boil at cert