The oxides of manganese are then filtered off, the filtrate neutralized with hydrochloric acid, and evaporated to about one fifth of its original volume. Strong hydrochloric acid is them added in excess, and on cooling the acid potassium salt of paranitroorthosulphobenzoic acid separates in very slender colorless needles completely filling the liquid. For the success of this operation it is important that the potassium salt of paranitroorthotoluenesulphonic acid and the potassium hydroxide should both be perfectly dissolved before they are heated together. If the two substances lie together in solid form at the bottom of the flask, a very slight elevation of temperature leads to the formation of an extremely troublesome red substance, which is very difficult to remove. It is almost impossible to remove it from the oxidation product by recrystallization, since any considerable amount of it has a marked influence on the solubility of the salt, rendering it much more soluble. It persists throughout all subsequent transformations of paranitroorthosulphobenzoic acid, and should therefore be carefully avoided. Otto Fischer[11] has shown that in concentrated solution, potassium hydroxide acts on nitro derivatives of toluene, with the formation of various colored substances derived from stilbene. In the case of paranitroorthotoluenesulphonic acid, he describes the substance formed as possessing a cherry red color. The reactions involved in its formation are: CH₃ HC ============ CH / / \ 2C₆H₃——SO₂OK = C₆H₃——SO₂OK KO.O₂S——C₆H₃ + 2H₂O \ \ / NO₂ \ / \ /-- O --\ / N N \-- O --/ By oxidation this passes to a nitro compound of the composition HC============ CH / \ C₆H₃——SO₂OK KOO₂S——C₆H₃ \ / NO₂ O₂N It was no doubt the formation of substances of this nature that occasioned the color observed in some of the oxidations. [11] Ber. XXVI-2231; XXVIII-2281 The only effective method of separating this colored substance was found to be to pass to the neutral salt of paranitroorthosulphobenzoic acid, by making the solution slightly alkaline. The salt of this colored substance is also formed and the two can be separated by a few recrystallizations in a fairly satisfactory manner. The yield in both of the transformations involved in the preparation of paranitroorthosulphobenzoic acid does not fall far short of the theoretical. III. Preparation of the Symmetrical Chloride of Paranitroorthosulphobenzoic Acid.