Xylidine scarlet, although at the time of its introduction the only true azo-scarlet likely to come into competition with cochineal, was still somewhat on the orange side. The cumidine dye would obviously be nearer the desired shade. To meet this want, cumidine had to be made on a large scale, but here practical difficulties interposed themselves. The quantity of cumene in the light oil is but small, and it is associated with other hydrocarbons which are impurities from the present point of view, and from which it is separated only with difficulty. A new source of the base had therefore to be sought, and here again we find chemical science ministering to the wants of the technologist. It was explained in the last chapter that aniline and similar bases can be methylated by heating their dry salts with methyl alcohol under pressure. In this way dimethylaniline is made, and dimethyltoluidine or dimethylxylidine can similarly be prepared. Now it was shown by Hofmann in 1871, that if this operation is conducted at a very high temperature, and under very great pressure, the methyl-alcohol residue, _i.e._ the methyl-group, does not replace the amidic hydrogen or the hydrogen of the ammonia residue, but the methylation takes place in another way, resulting in the formation of a higher homologue of the base started with. For example, by heating aniline salt and pure wood-spirit to a temperature considerably above that necessary for producing dimethylaniline, toluidine is formed. In a similar way, by heating xylidine hydrochloride and methyl alcohol for some time in a closed vessel at about 300 deg. C. cumidine is produced. Hofmann's discovery was thus utilized in 1882, and by its means the base was manufactured, and cumidine scarlet, very similar in shade to cochineal, became an article of commerce.