EXAMINATION OF GRAHAM FLOUR. 41 



mills. The gliaclin ratio of the coarse middlings is very low. A 

 macroscopic examination of this product shows it to contain a con- 

 siderable amount of bran. 



No. 9121 is made from the flour of the fifth break obtained in the 

 regular run of milling white flour. To this product bran is added. 

 The flour used represents about an 8 per cent second clear. The 

 results of the separation on the sieves show that the flour must 

 have been made from a mixture of flour with bran, because only 

 0.6 per cent of coarse middlings was found in the sample and less 

 than 10 per cent of fine middlings. The sample shows approxi- 

 mately the normal amount of bran and shorts, and about the proper 

 amount of ash, but the ash content of the flour passing through the 

 109 sieve, namely 0.87 of 1 per cent, indicates more or less a low- 

 grade product. 



No. 9175 is made by mixing several products while in the process 

 of making ordinary flour, namely, the fifth and sixth middlings 

 before reduction, the first, fourth, and fifth break flours, and the 

 flour from germ middlings. The miller, however, does not know 

 in what amounts these are mixed, as he neither weighs nor measures 

 them. He depends upon the appearance of the resulting product. 

 The percentage of ash and pentosans in the original sample indicates 

 that some of the bran has been eliminated in the preparation of this 

 flour. 



No. 9217 is a mixture of 40 per cent patent flour, 15 per cent 

 small bran, shorts, and tailings, and 45 per cent of low-grade flour. 

 The patent flour used is a 70 per cent patent made from a soft winter 

 wheat. One to two per cent of the coarse bran is taken off. The 

 amount of bran is somewhat low, otherwise the sample analyzed 

 similar to normal Graham flour with the possible exception that the 

 ash content of the material passing through the 109 sieve is a little 

 above the average for a soft winter-wheat flour. This amount of 

 ash, namely, 0.84 per cent, is about equal to that obtained from a 

 mixture of patent flour and a low-grade flour. 



The percentage of bran or material left on No. 20 sieve varies from 

 to 14.6, of shorts on No. 40 sieve from 3.8 to 14.5, of combined bran 

 and shorts from 8.9 to 21.4. The coarse middlings showed a varia- 

 tion of 0.6 to 15.0, while the fine middlings varied from 5.2 to 30 

 per cent. The combined middlings varied from 7.8 to 42.9. The 

 amount of flour passing the 109 sieve varied from 46.0 to 82.4, half 

 of the samples yielding over 60 per cent of this product. The per- 

 centages of ash, fiber, and pentosans averaged, as a rule, lower than 

 those of the same constituents in Graham flour. In fact the ash 

 of these samples is so low as to lead to the conclusion that most of 

 them are more of the nature of so-called wheat meal than of imitation 

 Graham flours. Of 63 samples of soft wheats analyzed for ash, 



