192 Sanitary Studies of Baking P[Owders 



DiscussioN BY L. J. CuRTMAN AND P. Gross. In the A.O.A.C. 

 method (unmodified by Steel) the weighed precipitate may be regarded 

 as consisting of ALOs-Fe^Og-A'PoOg. The precipitate is analyzed for 

 FcgOg and P^Og. The difference between the total weight of the 

 precipitate, and the sum of the FeoOs and PsOg separately determined, 

 will give the AUOs in the weighed precipitate regardless of whether 

 the AIPO4 or FePO^ was partially decomposed in the process of wash- 

 ing or ignition. Barring the errors inherent in the separate determina- 

 tions of the Fe^Og and PoOg, this method will give a correct value for 

 the AI0O3 in the precipitate. 



Steel's method can be valid, from purely theoretical considerations, 

 only on the assumption that the PaOg in the weighed precipitate is just 

 sufificient to form the normal phosphates with the AI and Fe present ; 

 in other words, that the precipitate consists of a mixture of pure 

 FePO^ and AIPO4. The results given in our Tables 2 and 3 (pages 

 168-9) show conclusively, however, that the above assumption (tacitly 

 made by Steel), as to the composition of the precipitate, is unwar- 

 -ranted. It is quite conceivable that the error in Steel's method, as 

 noted in our Table 2, is unduly magnified by the relatively large 

 amounts of Fe and AI involved, and that with smaller quantities (such 

 as were used by Howe and by Steel) the error would be less apparent. 

 Experiment, only, can decide this point. We feel, however, that the 

 method of Steel, if reliable, should show a better agreement with the 

 theory than found. The addition of insufficient sodium phosphate 

 cannot be held responsible for the results in Table 2 (page 168), for 

 the reason that in these, as in all our determinations, special care was 

 taken to add a quantity of phosphate in excess of that required to 

 combine with all the AI and Fe present. 



It is regretable that neither Howe nor Steel deemed it necessary to 

 standardize their aluminum sol's by methods other than those in which 

 the AI is precipitated as phosphate. Since the object of this investiga- 

 tion was the comparison of methods finally depending upon the pre- 

 cipitation of AI as phosphate, it would seem imperative that the stand- 

 ardization of the aluminum sol. be based on a principle other than one 

 used in the comparison of the methods. Equally regretable is the 

 failure of Howe to supply details as to the manner in which he 

 standardized his iron sol. This is a serious Omission in view of the 

 important röle which the iron plays in Steel's method. 



An interesting confirmation of our results is to be found in some of 



