i88 Sanitary Studies of Baking Powders 



The method of Schmidt and Hoagland gave more satisfactory 

 results as may be seen by an examination of Table 2. There seems 

 to be a tendency f or the results to be slightly higher than the theo- 

 retical, which is particularly noticeable in Det'ns 31 and 32. These 

 determinations were made upon samples of gastric juice to which a 

 known amount of the Standard sol. of alum had been added, and, as 

 the material was not ashed previous to the analysis, there was more 

 or less organic matter present, which may explain the high values 

 obtained; although the results of Det'ns 35 and 36, which were car- 

 ried out upon ashed gastric juice to which a known amount of the 

 Standard alum sol. had been added, are nearly as high proportion- 

 ately as those from the unashed juice. The absolute error in Det'ns 

 35 and 36 is, however, much lower than the error in Det'ns 31 

 and 32. 



The determinatlon in blood is not as accurate as is to be desired, 

 yet the values obtained in Det'ns 39, 40, 43, and 44, are in much 

 closer agreement with the theoretical than those obtained by the 

 method of Steel (and Kahn). 



General conclusions. i. In our hands the method of Schmidt 

 and Hoagland proved more satisfactory than the method of Steel 

 (modified by Kahn), for the determination of aluminium in blood 

 and gastric juice, to which known amounts of aluminium had been 

 added. 



2. It is not feasible to determine aluminium (by the Schmidt- 

 Hoagland method) in gastric juice, without first ashing the material. 



