EXAMINATION" OF GRAHAM FLOUK. 29 



cent, the shorts from 4.9 to 11.8. These differences are due, of 

 course, to the thoroughness or fineness of grinding. In the same 

 way the amount of middlings will show a variation, the table showing 

 that the percentage varies from 29.1 to 63.6, while the flour which 

 passes through the 109 sieve shows a variation of from 24.9 to 53.4 per 

 cent. In no case does the amount of flour which passes through the 

 109 sieve exceed 55 per cent. Generally in samples of true Graham 

 flours the percentage of nitrogen is greatest in the bran and very 

 slightly less in the shorts, while it is considerably less in the middlings 

 and still less in the flour passing through the 109 sieve, so there is a 

 gradual gradation in the amount of nitrogen from bran to flour. 

 In every case the percentage of nitrogen in the flour of the samples 

 in Table 2 is less than that found in the bran or shorts, whereas the 

 contrary very often happens in the case of imitation Graham flours, 

 as will be seen in Table 5. 



The low amount of bran found on the No. 20 sieve, especially in 

 Nos. 6636, 6951, 6952, and 6953, is probably due to the fact that the 

 wheat from which these samples were milled was of the hard winter 

 variety. As a general rule Graham flour is made from soft winter 

 wheats, in which case the amount of bran is likely to be greater on 

 account of the fact that it is not so finely ground in the process of 

 milling. 



In looking over the gliadin ratio of the various products separated by 

 the sieves it is seen that there is a gradual increase from bran to flour. 

 The bran and shorts, however, are very much alike. They contain 

 practically the same amount of nitrogen and the same amount of 

 alcohol-soluble nitrogen, thus giving almost identical gliadin ratios. 

 The middlings, however, show a much higher gliadin ratio. Gener- 

 ally the higher the gliadin ratio of the middlings the less bran is found 

 therein; in other words, the purer the middlings are. The same 

 statement applies to the flour passing the 109 sieve. If we assume 

 that the gliadin ratio of the middlings is 100 the average of that of 

 the flour passing the 109 sieve is about 140. (See Table 11, p. 21.) 



Considering these samples as a whole neither the data on the phys- 

 ical separation nor the nitrogen and gliadin ratio of the flour passing 

 the 109 sieve would indicate that they were anything but true Gra- 

 ham flour, except possibly the low amount of bran and shorts found 

 in four of the samples. 



In sample No. 6951 the amount of flour passing the 109 sieve is 

 relatively small. The fact that hard winter wheat was used in the 

 preparation of this sample might lead one to expect that break flour 

 would form a large part of this product. The consequence is that 

 the percentage of nitrogen in this product is about the same as the 

 percentage of nitrogen in the middlings obtained from the same flour, 

 while the gliadin number of this product is the lowest of the 12 



