EXPERIMENT STATION BULLETINS. 611 



does give fairly satisfactory results. The alcoholic calcium chloride 

 method is much its inferior, both in convenience and accuracy and pos- 

 sesses, in addition, certain theoretical defects which may possibly in- 

 troduce significant errors. These are (1) the assumption that the acid 

 or alkali required to produce neutrality in the solution of ammonium 

 chloride and hydrochloric acid (or ammonium hydroxide) is the same 

 as that required to give a neutral solution of ammonium citrate and 

 citric acid (or ammonium hydroxide) and (2) the fact that a solution 

 neutral to cochineal is not neutral to all other indicators nor is it 

 neutral in the truly physical-chemical sense. This latter point has been 

 made by McCandless (38) who also found difficulty in the use of this 

 method. 



McCandless' (37) (38) analytical method might be successfully 

 adapted to the purpose of securing strictly neutral solutions within the 

 limits of error of the determinations involved, could some reliable method 

 be found for getting the composition of a truly neutral solution. 



Such a method is in fact now available in the electrometric one of 

 Hildebrand (27). By this method it is possible to prepare a neutral 

 solution by determining electrometrically the concentrations of the 

 hydrogen and hydroxyl ions. It necessitates apparatus and technic 

 too involved for the average technician, which fact will probably always 

 prevent its widespread adoption. A procedure will however be pro- 

 posed later which it is believed will be exempt from some of these dis- 

 advantages. 



In this same class of methods fall those whereby neutrality is sought 

 through some means of self adjustment of the solution such as the 

 removal of excess ammonia by heating as proposed by Herzfeld and 

 Feuerlein (2G) and Cook (11) or by aeration as suggested by Washburn 

 (52). The fallacy of the principle of this method has been pointed 

 out by Eastman and Hildebrand (15) and the actual uncertainty of 

 the results obtained has been shown by McCandless (38). In such a 

 procedure there is always danger that the removal of ammonia will not 

 stop at the right point but will proceed beyond it and leave the solution 

 acid, a thing which it has been claimed actually does take place in 

 determinations made at too high temperatures with solutions initially 

 neutral or alJcaline unless suitable precautions are taken to prevent it.^ 



(b) Metliods for the preparation of solutions of triammonium citrate: 

 The end sought in this class of methods is attainable by more ways than 

 in the former one. Chronologically, McCandless' (37) analytical 

 method was the first., By it one can, of course, prepare within the 

 limits of accuracy of the analytical procedures involved, a solution in 

 which the ratio of the two ingredients, citric acid and ammonia, is that 

 in which they occur in the normal salt. There is involved neither un- 

 usual technic nor expensive apparatus such as is required for the 

 physical chemical methods. It approximates in time and effort the 

 preparation of any standard solution which must be checked analy- 

 tically. 



The various physical chemical methods, i. e., the electrical conductivity 

 method (23) (42), the temperature method of Bell and Co well (6), 

 and the electrometric one of Hildebrand (27), while perhaps more or 



'Two solutions carefully prepared in thi.^ laboratory according to the directions of Cook (11) and of 

 Washburn (52) were found to have reactions corresponding respectively.to Ph 5.7 and 5.5. 



