The details of the method are as follows:  A solution of manganous sulphate slightly stronger than No. 1 is prepared.

The difference between 100 c.c. of it and 100 c.c. of No. 1 is ascertained, according to the method of Volhard, by means of solution No. 3.

The contents of E, E, together with the rinsings from the tubes, are poured into a capacious flask. 100 c.c. of the manganous sulphate and a few drops of nitric acid are then added, and the whole boiled.  Finally, the excess of manganous sulphate is determined, in the manner described by Volhard, by means of solution No. 3.  Subtracting from the total amount of permanganate thus used the quantity required to equalize the 100 c.c. of solution No. 1 and the 100 c.c. of the manganous sulphate, we shall have the quantity of permanganate reduced by the nitric oxide.

It must, however, be remembered that the value of solution No. 3 is now to be calculated on the basis of the equation KMnO_{2} + NO = KNO_{3} + MnO_{2}.  One molecule of permanganate equals one molecule of nitric oxide when manganous sulphate is used, since no part of the permanganate employed in this method is reduced below the superoxide condition.  In other words, solution No. 3 now represents only three-fifths as much nitric acid as it does when oxalic acid is used.

The results obtained by this method were moderately satisfactory, but not quite so exact as those obtained when oxalic acid was used.  A series of four determinations gave differences, between the volumes of permanganate calculated and used, of 0.05 to 0.15 c.c.

The principal objection to the method lies in the difficulty of determining, in the presence of the brown oxide of manganese, the exact point at which the oxidation is complete.

The carbon dioxide generator, A, was devised by us to take the place of the ordinary generators, in which marble is used.  We have found that a submersion of twenty hours in boiling water does not suffice to completely remove the air which, as is well known, is contained in ordinary marble; hence some other substance must be employed as a source of the gas.  In the apparatus which we are about to describe, the acid carbonate of sodium is used.

It consists of a long, narrow cylinder (450 x 60 mm.); a tightly fitting rubber stopper, through which three tubes pass, as shown in the figure; a small cylinder, F, containing mercury; and a sulphuric acid reservoir, G.

The tube, g, is drawn out to a fine point at the end and curved, so that the acid which is delivered into A falls upon and runs down the outside of the tube.  The tube, h, dips under the mercury in F. G and g are connected by means of a long piece of rubber tubing which is supplied with a screw pinch cock.

The apparatus is made to give any required pressure by raising or lowering G and F; but the elevation of G, as compared with that of F, should always be such that the gas will force its way through h rather than g.  The upper part of the cylinder, F, is filled with cotton wool to prevent loss of mercury by spattering.