When the content of manganese is large, it is sometimes rather long before the solution is ready for titration. The reason of this appears to be that a part of the manganese is first precipitated as hydrated sesquioxide, which is afterward oxidized to hydrated peroxide, for the upper portion of the liquid may sometimes be colored by chameleon, while the lower portion, which is in closer contact with the precipitate, is less colored or absolutely colorless. From this we also see how advisable it is to stir the liquid frequently during titration. Toward the close of it, it is also advantageous, when the contents of manganese are large, to warm the solution to about 50° C., because the removal of color is thereby hastened. When the fluid, which is well stirred after each addition of chameleon, has obtained from it a perceptible color, which does not disappear after several stirrings, the whole of the manganese is precipitated and the color of the solution remains almost unchanged after the lapse of at least twelve hours. When the content of manganese is large the solution may be divided into two equal portions, one of which is first to be roughly titrated to ascertain its content approximately, after which the whole is to be mixed together and the titration completed, which can thus be performed with greater speed and certainty. If too much chameleon has been added, one may titrate back with an accurately estimated solution of manganese, which is prepared most easily by evaporating fifteen cubic centimeters chameleon solution down to two or three cubic centimeters, boiling with two to three cubic centimeters hydrochloric acid so long as the smell of chlorine is observed, and then diluting the solution to ten cubic centimeters, when one cubic centimeter of it corresponds to the same measure of chameleon. With respect to the delay which must take place during the titration in order to give the precipitate time to fall, it is advantageous, in order to save time, to work with several samples; but it is, in such a case, desirable to have a separate burette for each sample, in order to avoid noting every addition of the chameleon solution and afterward adding them up. If burettes are wanting, and one must be used for several samples, a Mohr's burette with glass cock is the most convenient to use. For the titration of iron with chameleon solution, the latter is commonly used of such a strength that 0.01 gramme of iron corresponds to about one cubic centimeter of chameleon solution, which is obtained by dissolving 5.75 grammes permanganate of potash in 1,000 cubic centimeters water. The titration is determined by means of iron, a salt of iron or oxalic acid. A drop of such a solution, corresponding to about one-twentieth cubic centimeter, or 0.0001 gramme Mn, is sufficient to give a perceptible reddish color to 200 cubic centimeters of water. As what takes place in the titration of iron with chameleon is indicated by the following formula,