A new phase in the technology of tar-products was entered upon when Witt caused a diazo-salt to act upon diamidobenzene. This was the first industrial application of Griess's discovery. Azobenzene, which was discovered by Mitscherlich in 1834, and azotoluene are now manufactured by reducing nitrobenzene and nitrotoluene with mild reducing agents. These parent compounds are not in themselves colouring-matters, but they are transformed into bases which give rise to a splendid series of azo-dyes, as will be described subsequently. Let it be recorded here that these two compounds are to be added to the list of valuable products obtained from benzene and toluene. And it must also be remembered that the introduction of these azo-colours has necessitated the manufacture on a large scale of sodium nitrite as a source of nitrous acid. Without entering into unnecessary detail it may be stated broadly that this salt is made by fusing Chili saltpetre, which is the nitrate of sodium, with metallic lead, litharge or oxide of lead being obtained as a secondary product. Then again, the manufacture of Bismarck brown requires dinitrobenzene, this being made by the nitration of benzene beyond the stage of nitrobenzene. The brown is made by reducing the dinitrobenzene to diamidobenzene, and then treating a solution of the latter with sodium nitrite and an acid. The azo-colour is formed at once, and no special refrigeration is required in this particular case. It has already been stated that the old aniline yellow of 1863 is no longer used on account of its fugitive character. In 1878 Graessler found that by the action of very strong sulphuric acid this azo-compound could be converted into a sulpho-acid in just the same way that magenta can be converted into acid magenta. Under the name of "acid yellow" this sulpho-acid is now used, not only as a direct yellow colouring-matter, but as a starting-point in the manufacture of other azo-dyes. The use of acid yellow for this last purpose will be dealt with again in the next chapter.