174 C. L. BRIGHTMAN, MEACHEM AND ACREE 



trophotometer. The measurements are therefore more accurate 

 than ordinary colorimetric data. The actual data and calcula- 

 tions involved are of course very extensive. We shall not give 

 all these numerical values in this paper but state how the final 

 figures given in the tables were obtained. The method was to 

 measure the absorption for several wave lengths in the band. A 

 smooth curve was drawn with wave lengths as abscissae and ab- 

 sorption indices as ordinates. The ordinate with the maximum 

 value on this curve was taken as the absorption index for the mid- 

 dle of the green band. This index would change, of course, for 

 different depths of cell and for various concentrations of the indi- 

 cator. In order, therefore, that Birge's work might be utilized, 10 

 the value of the index was reduced, by Beer's law, to the value 

 it would have were the cell 6 cm. deep and the concentration of 

 the phenolsulfonphthalein N/1000. The values of the index 

 (N a ) thus obtained are given in table 1, column 3. 



Birge found in his work that, with different dilutions, the value 

 of the absorption index for the green band when the indicator was 

 wholly transformed showed some slight variation which will not 

 appreciably modify the results in this paper. The value given 

 for N/5000 concentration was 225. This was then taken pro- 

 visionally for our value of N but will be measured for each con- 

 centration of the different indicators and of the phosphate and 

 other buffers. Column 4 in the table gives the values of (1 — a) /a 

 as computed from the values (N — N a /N a . 



The hydrogen ion concentration (H) of each solution was meas- 

 ured by the Loomis-Acree method 11 involving the N/10 KC1- 

 HgCl electrode, 4.1NKC1 as a connection link and the hydrogen 

 electrode, and recorded in column 5 of the table. Column 6 

 gives the value of the ionization constant as computed from the 

 values of (1 — a) /a and (H). It will be seen that the value of K 



10 Unpublished work by Birge, Brightman, Meacham, Hopfield and Acree. 

 See also Birge and Acree: J. Am. Chem. Soc, 41, 1031, Brightman, Hopfield, 

 Meacham and Acree, Ibid., 40, 1940. 



11 Loomis and Acree: Am. Chem. J., 46, 585, 621, Myers and Acree: Ibid, 50, 

 396. Clark, Myers and Acree, J. Phys. Chem. 20, 243. The errors involved in 

 the Bjerrum method, and in the Walpole modification, which was used by Lubs 

 and Clark, will be discussed in another article. 



