electrode and the unknown solution. Electrical connection between the electrode and the potentiometer is made through platinum wire sealed into the bottom of the vessel and either wired directly, as. in D, or through a short column of mercury in the small glass tubes sealed at the bottom of the electrode (C and D). In the dipping electrodes, A and B, the calomel electrode proper is contained in the small inner tube, which, near the bottom and just above the calomel layer, has a small opening through which the electrode communicates with the potassium chloride solution contained in the outer jacket. This liimis jacket is closed by a ground joint which is moistened with potassium chloride solution which passes the current when immersed in the test solution. It is essential that pure materials be used in preparing the electrodes in order that they may have the accepted emf values. Mercury may be purified by allowing it to fall in a fine spray through diluted nitric acid. It is then washed with distilled water, dried, and distilled in vacuo. Reilly and Rae [4] give the following directions for purifying mercury by electrolysis: "The mercury is placed in a large dish inside which is a small dish containing a small quantity of mercury. Platinum-wire electrodes sealed into glass tubes make contact with the two portions of mercury, that in the smaller vessel being the cathode. An electrolyte, consisting of 90 parts of water, 5 parts of sulfuric acid, and 5 parts of nitric acid, is poured in so as to cover the upper edge of the smaller vessel to a depth of 2 cm. A current of 0.5 ampere is passed for 2 hours for each kilogram of mercury in the outer vessel. The mercury in the outer vessel is then removed, washed, dried, and distilled in vacuo." For the preparation of calomel a portion of the purified mercury is dissolved in pure nitric acid that has been redistilled and slightly diluted. The mercury nitrate solution is poured into a large excess of distilled water containing some nitric acid. To this solution is slowly added dilute hydrochloric acid purified by distilling pure 20percent acid, discarding the first and last portions of the distillate. The precipitate is repeatedly washed with distilled water, preferably by decantation. Some free mercury should be present throughout the process to prevent the formation of mercuric salt. At the completion of washing, the calomel should be intimately mixed with finely divided mercury by shaking. Half-cell I Half-cell III. Half-cell IV. Half-cell V. Half-cell VI. Half-cell VII. TABLE 34.—Arbitrarily standardized values for half-cells 1 (H+)=1|H2 (1 atm), Pt. KCl (sat.) KCl (0.1 N), HgCl|Hg. KCl (sat.), HgCl|Hg. KCl (sat.) HCI (0.1 NH (1 atm), Pt. 1 KCl (sat.) (KH phthalate (0.05 M)[H, (1 atm), Pt. KCl (sat.) Na acetate (0.1 M)| Ha(1 atm), Pt. (H+)=1, quinhydrone Pt. KCI (sat.) [HC1 (0.1 N), quinhydrone Pt. 1 W. M. Clark, The Determination of Hydrogen Ions, 3d. ed., p. 672 (Williams & Wilkins Co., Baltimore, Md., 1928). The calomel-mercury mixture, before being placed in the vessel, is repeatedly shaken with small quantities of the potassium chloride solution chosen for the cell. The potassium chloride solution used to complete the filling of the cell is also saturated with calomel. The calomel half-cell may be connected directly with the test solution or through a salt bridge formed usually by a saturated solution of potassium chloride flowing from a reservoir through a side arm of the vessel and making contact with the calomel-saturated potassium chloride of the cell. After a measurement, the solution at the end of the side arm may be flushed out by slightly turning the stopcock. In electrodes A and B (fig. 56) the larger outside tube contains saturated potassium chloride and the inner tube is the calomel electrode proper. Clark's [3] table of values for several half-cells commonly used in the determination of hydrogen-ion concentration is reproduced in table 34. It is pointed out by Clark that discrepancies may be found in certain values given in the table and that certain values are to be regarded as tentative. This applies particularly to the temperature error in the saturated calomel half-cell. "On the other hand, the potential of a cell composed of a hydrogen or quinhydrone half-cell and a saturated potassium chloride calomel half-cell has a small temperature coefficient. . . so that the temperature value may be in considerable error without causing great error in potential.” (c) HYDROGEN ELECTRODES A hydrogen electrode is formed by saturating platinum black (coated on platinum foil or wire for rigidity) with hydrogen gas. When such an electrode is placed in contact with a solution containing hydrogen ions, a difference of potential is established at the electrodesolution interface analogous to the potential difference between a metal electrode and a solution containing its ions. The emf of such a half-cell is not directly measurable, but if two half-cells be connected so that the electrolytes form a sharp liquid junction, the total emf of the concentration cell so formed may be measured potentiometrically. If the emf of one half-cell is known, that of the other may be computed. The emf of such a concentration cell, ignoring the liquid-junction potential, is represented by ᎡᎢ C E- In nF (80) where R is the gas constant=8.31507 volt-coulombs in absolute units; T is the absolute temperature=273.1+t° C; n is the valency of the electrode element; and F (the faraday) is the charge on 1 g equivalent of the ion=96,500 coulombs. C and C, are the ion concentrations in the two halves of the cell. Substituting and multiplying by 2.3026 to convert to common logarithms, we obtain T n (81) Since measurements of E are customarily made in terms of international volts instead of absolute units, the numerical value in eq 81 is converted by dividing by 1.00042 [3, p. 250], and we obtain T C E=0.000198322log C (82) For the hydrogen electrode, let us assume that the concentration in one of the half-cells is normal with respect to hydrogen ions (1.008) g/liter) and write C=1, and in the other half-cell let the hydrogen-ion concentration be unknown and write C1=[H+]. The valence, n=1, and eq 82 then becomes E=0.000198322T log [H+] 10g面 (83) 903232 O - 50 - 20 The normal hydrogen half-cell is not used as a reference standard in practical pH measurements. As a primary standard of reference, Clark [3, p. 480] recommends the 0. 1 N calomel half-cell, referred to the potential of the normal hydrogen half-cell as equal to zero, and specifies as follows: "It shall be assumed, arbitrarily, that in the cell Pt, H2 (1 atmos.) |H+ (activity, x)|KCÍ (sat.)|KCl (0.1 N), HgCl(sat.) [Hg the poB D A tential difference at B remains constant as a varies and that the sum of the potential differences at B, C, and D is as follows at each indicated temperature. [These values are given in table 34, column 2.] "The standard experimental meaning of pH shall be the potential of the above cell considered as of positive numerical value, less the above value for the calomel half-cell pertaining to the temperature used, the difference being divided by the numerical quantity 0.000198322T, where T is the absolute temperature." The statement of the last paragraph may be formulated in the equation This does not preclude the use of secondary standards, such as the saturated calomel half-cell or the quinhydrone electrode, but the "attempt shall be made to use this standard in accordance with the specifications made above." Hydrogen-electrode potentials vary with barometric pressure, and the apparatus should be so constructed that the gas pressure in the half-cell is the same as that of the surrounding atmosphere. In most routine measurements, a barometric correction may be omitted, but it should be included for exact measurements. The corrected value of pH is obtained by and the correction value E (bar.) is found by 1 12 Clark [3, p. 479] states: "Originally pH was defined by pH=log Actually the numerical values called pH have been determined by dividing the potential of a hydrogen cell by 2.3026RT/F. In the comparison of one solution with a standard solution of hydrion activity of unity, the rigid relation may be written where (H+) represents the hydrogen-ion activity of the solution under investigation. Consequently, the 1 " measured values called pH are log (H+)' where x is the value found by subtracting the pressure of aqueous vapor at the temperature, t° C, from the barometer reading. The hydrogen electrode is used for the standardization of buffer solutions and for checking methods in which other electrodes are used. It is useful in properly buffered solutions over a range of 0.0 to 14.0 pH. It is inaccurate in the presence of certain metals and dissolved gases and in the presence of sulfites or sulfurous acid. The coating of platinum black is sometimes clogged or "poisoned" and rendered inactive. The Clark rocking-electrode assembly, consisting of hydrogen and calomel electrodes, is shown in detail in figure 57, and mounted for use in figure 58 (E). The hydrogen electrode vessel, shown at E (fig. 57) is mounted in a clamp pivoted behind the rubber connection between J and H. This clamp runs in a groove of the eccentric, I, the rotation of which rocks the vessel. With the hydrogen electrode inserted at F, the vessel is filled as full as possible with water from D through cock C. The water is displaced with hydrogen through A after turning C to communicate with drain, B, and the vessel is flushed with succeeding changes of hydrogen. The vessel is rocked back toward C, which is closed, and the test solution is run in from reservoir, D, with G open toward B' until the vessel is half full. Cock G is now closed and cock C is opened so that hydrogen may enter through A to exert a continuous pressure on the liquid. The solution is then rocked until equilibrium is established with the liquid. The rocking alternately completely immerses and then exposes the electrode to the hydrogen gas. At M is a saturated potassium chloride-calomel electrode used as a working standard, and at N is a reservoir containing saturated potassium chloride used in making liquid junctions. At P is a battery of accurately made 0.1 N calomel electrodes used in standardizing the saturated calomel working cell. This battery may be connected with the system through the liquid junction at O by opening cock K. To measure the emf, the rocking is stopped, and E is opened so that a small amount of saturated potassium chloride may escape through |