[Vol. 6 



78 ANNALS OF THE MISSOURI BOTANICAL GARDEN 



range between P H 8.0 and 12.0. This same method was ex- 

 tensively employed with other indicators with an acid range 

 in the work with Dr. Duggar (Duggar and Dodge, '19), so that 

 it seems to have a wide application, although considerable 

 practice and an intimate knowledge of the minute color 

 changes of the indicator worked with is required before accu- 

 rate titrations, or rather series of titrations, can be made 

 rapidly. Five cubic centimeters of solution is a suitable 

 quantity to titrate, and the process of obtaining a graph is 

 essentially as follows : The cups are properly filled with col- 

 ored solutions, indicator, etc., then the solution is titrated by 

 the addition of a drop at a time until the first hydrogen ion 

 concentration is reached, in the present work P H 8.2. Then the 



standard color corresponding to P H 8.4 is placed in the cup 

 and the unknown titrated until this hydrogen ion concentra- 

 tion is reached. This is recorded and the process repeated 

 until the curve has been carried as far as desirable, or until 

 some disturbing factor enters which makes further titration 

 impossible. Below is a set of titrations taken at random to 

 show the accuracy of the method. Solution A was almost 

 black, solution B was a golden straw color, and solution C 

 was a tyrosin solution, consequently colorless. 



Later even a greater accuracy was obtained in the critical 

 region of P H 9.0 to 9.4, and I do not doubt but that greater 

 practice will enable one to refine the process still further. The 

 difficulty of estimating quantities of less than 0.01 cc, of 

 manipulating the stopcock, and of securing a burette tip which 



will drop less than 0.01 cc. without being easily broken, makes 

 the probable error of the single observation a better measure 

 of accuracy than percentage. 



N/50 Ba(OH) 2 was used as the alkali in titrating tyrosin, 

 since it is removed from harmful action other than lowering 

 the amount of free alkali in the solution because of its insol- 

 ubility, and also the presence of carbonate is more easily 

 detected. As soon as carbonate was noted, the solution was 

 discarded if in the burette, and filtered and restandardized 

 if in the stock bottle, although this precaution was probably 



unnecessary in view of the experimental error in other direc- 



