28 



GKAHAM FLOUR. 



Nos. 9940, 9946, and 9952 are made from durum wheat. The 

 amount of combined bran and shorts is greatest in No. 9940 and least 

 in No. 9952, the former having been passed three times through the 

 corrugated rolls and the latter having been ground between burrs. 

 The amount of flour is least in No. 9940 and greatest in No. 9952 and 

 so are the middlings. The nitrogen, ash, fiber, pentosans, and 

 gliadin ratio in all three samples agree quite well. As in the case of 

 hard winter wheat, the nitrogen of the flour passing through the 109 

 sieve in each case is equal to or greater than that of the fine middlings. 

 In No. 9952 the nitrogen of the flour is even greater than the nitrogen 

 of the bran. In each case the gliadin ratio increases from the bran 

 to the flour. 



The amount of middlings in No. 9940 is only 12.3, yet the fact that 

 the shorts are so high (56.4 per cent) indicates that a large proportion 

 of the middlings and flour particles are adhering to them. A small 

 amount of middlings considered by itself is not necessarily an indi- 

 cation that the flour is not genuine. 



Comparing all three sets of Graham flour produced from the three 

 varieties of wheat it is seen that the ash of the flour from the soft 

 winter wheat passing the 109 sieve is the lowest, and that from durum 

 wheat is highest. Yet the ash of the Graham flour from the soft 

 winter wheat is higher than that from the durum. This is explained 

 by assuming that in the process of grinding soft wheat there is a more 

 complete separation than when hard wheat is ground. In the latter 

 case the particles of bran are more easily broken up and they find 



their way into the middlings and into the flour. 







SAMPLES OF TRUE AND IMITATION GRAHAM FLOURS (TABLES 2, 3, 4, AND 5) PURCHASED 



ON THE MARKET. 



The samples in Tables 2 and 3 will next be described and the 

 results discussed. These 19 samples, as has already been stated, 

 were bought on the market. Twelve samples are true Graham flours, 

 while seven of them are imitation Graham flours, one of which was 

 labeled "Graham flour style." These samples, as well as those 

 reported in Tables 4 and 5, to be discussed later, were subjected to a 

 physical separation on sieves Nos. 20, 30, and 109, No. 70 not being 

 used. The material on the No. 20 sieve was called bran, that on the 

 No. 30 shorts, and that on the No. 109 middlings, there being no 

 distinction made between the coarse and fine middlings as was done 

 with the other samples. The material going through the 109 sieve 

 was called flour. 



Table 2 shows the variations in the percentage of these various 

 constituents in each sample, and likewise the percentage of nitrogen 

 and the gliadin number of the Graham flour and of each portion 

 separated. The bran content is seen to vary from 0.4 to 18.8 per 



