172 Sanitary S tu dies of Baking Powders 



with greater force, although the method has a correct theoretical 

 basis. 



Conclusions. I. The method of Steel has been shown to be 

 unreliable, due to the instabihty of FeP04. For the determination 

 of small quantities of aluminum in the presence of relatively large 

 amounts of iron, the method is unsatisfactory owing to the increas- 

 ing error occasioned by the latter. 



II. Accurate results were obtained by the method of Schmidt 

 and Hoagland both in pure sol. and in blood. 



III. From the point of view of technique, the method of Schmidt 

 and Hoagland is superior to that of Steel for the following reasons : 



(a) A direct gravimetric determination of aluminum is effected. 

 No Volumetrie sol. or Operations are required and the aluminum is 

 not found by difference (thus taking the sum of the errors). 



(&) The determination is made on the entire sample, not on an 

 aliquot portion of the sol. as in Steel's method and, as a consequence, 

 the error in dealing with small amounts is thus materially decreased. 



(c) The tedium of washing the precipitates is to a great extent 

 avoided without any sacrifice of accuracy. 



