\ LIBRARY OF CONGRESS 014 338 319 2 , TX 595 .F5 K4 Copy 1 UNITED STATES DEPARTMENT OF AGRICULTURE BULLETIN No* 839 Contribution from the Bureau of Chemistry CARL L. ALSBERG, Chief Washington, D. C. April 23, 1920 THE MICROSCOPICAL EXAMINATION OF FLOUR By GEORGE L. KEENAN, Microanalyst, and MARY A. LYONS, Microanalyst, Microchemical Laboratory CONTENTS Review of Literature Purpose of Investigation . . , Microscopical Method . . . , Sources of Variation in Method , Page I 3 8 5 Exmmination of Mill Stoclcs II Page Examination of Commercial Grades of Flour 16 Examination of Experimental Series of Flour 29 Summary 31 Biblioeraphy S2 WASHINGTON GOVERNMENT PRINTING OFFICE 1920 L^ U:\ D* or -a UNITED STATES DEPARTMENT OF AGRICULTURE BULLETIN No. 839 i i jrL^'^'«-rt. Contribution from the Bureau of Chemistry CARL L. ALSBERG, Chief jru^'^o**^ Washington, D. C. April 23, 1920 MICROSCOPICAL EXAMINATION OF FLOUR. By George L. Keenan, Microanalyst, and Mary A. Lyons, Microanalyst, Microchemical Laboratory. CONTENTS. Review of literature 1 Purpose of investigation 3 Microscopical method 3 Sources of variation in method 5 Examination of mill stocks 11 Examination of commercial grades of flour ... 16 Examination of experimental series of flour. 29 Summary 31 Bibliography 32 REVIEW OF LITERATURE. A review of the literature has shown very few methods for tlie microscopical examination of flours. In the great majority of the methods found, suggestions are offered for the separation of the wheat tissues from the starch material and the subsequent examina- tion of the offal under the microscope. The results obtained from such microscopical examination, however, are only roughly indicative of the offal that may be present. The work of Delaye (5)^ was concerned largely with the detection of foreign spores in flour and also with the presence of ergot. Girard (7) suggested the separation of the gluten from the starch and im- purities by forming the flour into a cake and washing it with running water. The starch and impurities were separated with a fine sieve, and the offal particles examined under the microscope. Kraemer dl) has offered a quantitative method for the examination of com- mercial flours by means of the microscope, this quantitative method to be preceded by a general qualitative examination. A small por- tion of the flour was weighed out, a few drops of a reagent added, and the number of typical starch grains or characteristic tissues enumerated in examining five different portions of the microscop- ical mount. Standard samples were employed for purposes of com- ' The numbers in parenthesis refer to the bibliography on page 32. 152332°— 20— Bull. 839 1 2 BULLETIN 839, U. S. DEPARTMENT OF AGRICULTURE. parison. As a rule, not less than 12 microscopical mounts were made of the standard and of the sample under examination. Kohn (10) weighed out one-half gram of the flour, and added 10 cubic centimeters of ether, shaking the mixture, to isolate the hairs and bran tissues which were subsequently identified under the microscope. Dedrick (4) placed the flour in question upon a glass, and exam- ined it for offal particles, either with the naked eye or by means of a powerful magnifier. Particles of bran, germ, or other impurities or substances differing from flour were enumerated, five and six trials being made and an average struck. In this way he attempted to differentiate between the so-called patent, straight, clear, break, and low-grade flours. Collin (3) considered the microscopical examination of flour quite extensively, although he did not take up the question from the stand- point of determining the relative amount of offal material present. The histology of the wheat grain is fully discussed and illustrated with a number of figures. Von Liebermann and Andriska (22) suggested a method for esti- mating the quality of wheat flour which might possibly be correlated with a microscopical examination. The quality of the flour with respect to the quantity of bran substance present was ascertamed by shaking the flour with chloroform, and observing the color of the portion which floated on the surface. The test might be rendered quantitative in the following manner: One gram of the flom* was shaken in a tube with 10 cubic centimeters of chloroform, and the mixture allowed to stand for one hour. The depth of color of the layer which then formed on the surface of the chloroform was com- pared with the colors of the layers produced when mixtures of finest white flours and variable portions of bran were subjected to similar treatment. These mixtures might contain quantities of washed bran ranging from to 2 per cent. The colors of the layers were to be observed from above. In connection with the work done by Moore and Wilson (15), Pat- terson has made a microscopical examination of the flour streams from the different machines of the mill, these streams being blended to form various finished fiours. Finished flours were also examined. His method consisted in weighuig out 3 milligrams (0.003 gram) of flour, dividing this into five portions on as many microscopic slides, wetting with water, covering with cover slips, and then counting imder the microscope the number of hairs and epicarp and seed- coat particles in the five slides. His results tended to show how these particles increased in number in streams from the lower-grade machines and were practically absent from those from the "top" of the mill. MICROSCOPICAL EXAMINATION OF FLOUR. 6 PURPOSE OF INVESTIGATION. From the review of the literature, it is apparent that heretofore the purpose of the microscopical examination of flour has usually been to determine the presence of adulterants, such as other flours, or even starches, spores, etc. The paper in which Patterson indi- cated the possibilities of an estimation of the offal content of a flour microscopically (15) suggested the work here reported. MICROSCOPICAL METHOD. For convenience, the microscopical method employed in this inves- tigation will be described under the following headings: (1) Appa- ratus, (2) technique, and ^3) counting bran particles and hairs. APPARATUS. 1. Microscopic slide with a ruled area about 22 millimeters square. The lines, which it is convenient to have about h millimeter apart, are ruled across the short diameter of the slide. 2. Cover glasses 22 millimeters square. 3. Compound microscope, with compensating ocular 12 X and 16 mm. apochromatic objective. 4. Scalpel, preparation needles, camel's-hair brush, spatula, alco- hol lamp, mechanical stage. 5. Assay balance. 6. Chloral hydrate solution about 1:1; preferably not any more concentrated. TECHNIQUE. Before undertaking the examination of a flour microscopically, the sample should be thoroughly mixed, and a composite sample with- drawn from various parts of the material. A 5-milligram portion of flour is carefully weighed out upon accurate balances, and the weighed portion transferred to the center of the ruled area on the microscopic slide. The scalpel is employed in removing flour from the weigh- ing pan to the slide, the small amount which can not be thus re- moved being easily brushed onto the slide with the camel's-hair brush. The flour being transferred to the slide, about 3 or 4 drops of chloral hydrate solution are mixed with the flour by means of the preparation needle. Add only enough chloral hydrate solution to fill the space be- neath the cover glass. The proper amount is usually about 4 drops when a pipette with a 1 -millimeter bore is employed. A pipette of larger bore releases too much solution at a time and is less convenient to control. It is important that the material be evenly distributed in the solution; otherwise flocculation of the flour will occur, rendering counting more difficult and less accurate. The square cover glass is next applied, and the slide heated over the alcohol flame until the 4 BULLETIN 839, U. S. DEPARTMENT OF AGRICULTURE. starch grains are dissolved, or the preparation "cleared," as is com- monly stated. Vigorous heating of the shde is to be avoided in order to prevent burning of the material before the preparation has been sufficiently cleared. After gentle heating, the slide is quickly trans- ferred to the stage of the microscope, where it is allowed to remain a short time before counting is begun. The cold stage causes the larger part of the air bubbles that may be present in the preparation to dis- appear, the very few that remain not hindering in the enumeration of the bran particles and hairs. Careful adherence to the details of this technique is necessary to insure a suitable slide for counting. If a slide is improperly prepared, the resulting count probably wiU not be representative of the flour under consideration. COUNTING BRAN PARTICLES AND HAIRS. A thorough acquaintance with the histology of the wheat grain is essential before attempting an examination of flours. An}^ standard work on microscopy or plant anatomy of the common food products contains adequate descriptions of the tissues of the wheat berry in various sections. The following brief description of the anatomy of the wheat berry ^ is given for the purpose of indicating the tissues which are depended upon for judging a flour with respect to its offal content. The wheat grain is, botanicaUy, the fruit of various subspecies and varieties of the genus Triticum. This grain or fruit consists of a series of tissue systems, the outermost of which is the pericarp which is composed of three layers, the epicarp, mesocarp, and endocarp. The pericarp is essentially the fruit coat or matured ovary wall. Within the pericarp is the testa (or spermoderm), rather yellowish- brown in color, and easily distinguished in either cross or surface sections under the microscope. Within the testa is a layer of rectan- gular cells (in transverse section) known as the aleurone layer, contain- ing protein material but no starch. This is essentially the outer layer of the endosperm or albumen of the seed. The remainder of the grain within the aleurone layer consists of very thin-walled parenchymatous cells packed full of starch grains. The small embryo, or germ, is located at the end opposite the bearded apex. A crease or groove passes longitudinally from the base of the grain to the apex. The essential purpose of milling is to produce the finely ground endosperm or starchy portion of the wheat grain as free as possible from bran particles, hairs, and germ tissues. These bran particles, hairs, and germ tissues are known as offal in miUing terminology. The wheat offal, therefore, consists primarily of all the tissue elements of • A. I'. Winton. The Microscopy of Vegetable Foods, 2(1 ed., pp. 65-73. 1916. MICROSCOPICAL EXAMINATION OF FLOUR. 5 the grain from, and including, the aleurone layer outward, and also the germ tissues. Botanically, the bran consists of the pericarp, or fruit coat, and the aleurone layer. In order to discover any relation that might exist between the bran particles and hairs and the various so-called gi-ades of flour, the microscopical method abeady partially described (page 3) was em- ployed to determine the number of bran particles and hairs ordinarily found, in varying amounts, in different classes of flours. This enu- meration consisted in methodically examining and recording all of the bran particles and hairs contaiued in any given slide. It is well to form the habit of always starting at the same point in the mount, as, for example, the lower right-hand corner of the sHde. The sUde is slowly moved by means of the mechanical stage, and aU of the bran particles and hairs detected outside the edge of the cover sUp counted. Each particle of spermoderm (with accompanying aleurone layer, if present), epicarp, cross-cell and intermediate-cell tissues, and hairs are given a value of one, no matter how small the particle or hair fragment may be, surface as weU as transverse sections being included. After the region outside the cover slip is carefully scrutinized, the slide is moved over the width of the space between the ruled lines, and another strip of the mount examined and the offal* counted. A bran particle with hairs attached is counted as so many hairs instead of being recorded, for the sake of convention, with the bran particle count. Germ tissues were not enumerated. This procedure, as de- scribed, is methodically followed until the entire slide has been examined. SOURCES OF VARIATION IN METHOD. In order to study the reliability of the method aside from its practi- cal apphcation to the examination of flour, a large niunber of tests were made having for their principal purpose the determination of the probable sources of variation and their extent. In considering this question it was recognized that there might be a variation due to one or aU of the following factors: (1) Personal equation, including one analyst's variation in counting the same slide on different days and the variation between two analysts counting the same slide on the same day; (2) daily variation due to the condition of light, etc.; (3) slide variation due to limits of accurate weighing of the test portion of flour; and (4) the variation in homogeneity of the bulk sample. 1 For the purpose of this investigation bran particles and hairs were considered as constituting tlie offal. 6 BULLETIN 839, U. S. DEPARTMENT OF AGRICULTURE. PERSONAL EQUATION VARIATION. COUNTING THE SAME SLIDES ON DIFFERENT DAYS BY ONE ANALYST. Table 1 gives actual data obtained from counts made by each of two analysts working upon three slides which were prepared from the same bulk sample and upon which they made two counts on each of three successive days. Table 1. — Results of counts of same slides by two analysts on different days. Date. SUde. Analyst. Count No. Bran particles. Bairs. Total 1918. Jan. 7 A A B B C c a a B B C C A A B B C C A A B B C C A A B B C C A A B B C C Keenan 1 2 1 2 1 2 1 2 1 2 1 f 2 1 2 1 2 1 2 1 2 1 2 1 2 1 2 1 2 1 2 1 2 1 2 87 92 60 60 87 81 103 114 86 80 90 87 76 82 60 49 62 64 100 % 85 77 89 83 80 83 49 53 60 77 104 100 78 77 86 86 59 64 60 58 62 74 58 52 64 58 62 57 66 69 55 48 60 68 54 56 52 55 65 59 65 66 61 57 66 71 54 55 55 55 65 62 146 Do.. do 156 Do.. do 120 Do.. do 118 Do.. ....do 149 Do.. do 155 Do.. Lyons 161 Do.. do 166 Do.. do 150 Do.. .... do 138 Do.. do 152 Do.. do 144 Jan. 8 Keenan 142 Do.. do 151 Do.. do 115 Do.. do 97 Do.. do 128 Do.. .. do 132 Do.. Lyons 154 Do . . do 152 Do.. do 137 Do.. do 132 Do . do 154 Do.. do 142 Jan. 9 Keenan . . 145 Do.. do 149 Do.. do 110 Do. do 110 Do.. do 126 Do . do 148 Do Lyons 158 Do.. do 161 Do.. do 133 Do .do... 132 Do. do 151 Do.. do 148 For the purpose of emphasizing certain salient points, the results recorded in Table 1 have been rearranged in Table 2, in considering which it is necessary to regard the different portions carefully. Keenan's greatest variation in two counts of bran particles on a given slide on any one day was 17 points (slide C, Jan. 9, 1918), while Lyons' greatest variation was 11 points (slide A, Jan. 7, 1918). In the matter of counting hairs the greatest variation in the counts obtained on a given slide on any one day by Keenan was 12 points (slide C, Jan. 7, 1918), while Lyons' greatest similar variation was 6 (in several instances). In these cases it appears therefore that the personal variation due to the error of counting probably would not exceed 17 points in the case of particles or 12 points in the case of hairs. MICROSCOPICAL EXAMINATION OF FLOUR. Table 2. — Variation in counting of each analyst. SUde. Count No. Bran particles. Hairs. Jan. 7, 1918. Jan. 8, 1918. Jan. 9, 1918. Jan. 7, 1918. Jan. 8, 1918. Jan. 9, 1918. A Keenan. 1 2 1 2 1 2 Lyons, I 2 I 1 2 87 92 60 60 87 81 103 114 86 80 90 87 76 82 60 49 62 64 100 96 85 77 89 83 80 83 49 53 60 77 104 106 78 77 86 86 59 64 60 58 62 74 58 52 64 58 62 69 66 69 55 48 66 68 54 56 52 55 65 62 65 A 66 B 61 B 57 C 66 C . 71 A 54 A... 5S B 5i B... 55 C 68 C 63 COUNTING THE SAME SUDE ON THE SAME DAY BY TWO ANALYSTS. The variation between the counts made by two analysts on the same slide on the same day is demonstrated by comparing the daily averages * obtained by each of the two analysts. These data are compiled in Table 3. Table 3. — Varintion in counting of two analysts on same day. Date. Analyst. Bran particles. Hairs. Slide A. Slide B. Slide C. Slide A. Slide B. Slide C. 1918. Keenan 89 108 19 79 98 19 81 105 24 60 83 23 54 81 27 51 77 26 84 88 4 63 86 23 68 86 18 61 55 6 67 55 12 65 54 11 59 61 2 51 53 2 59 55 4 6S Jan. 7 • Lyons 59 Variation 9 Keenan 67 Jan. 8 ■ Lyons 68 Variation 5 Keenan... 68 Jan. 9 ■{Lyons 61 Variation 5 The table shows an average variation in the comit of bran par- ticles of 20, with a range of from 4 to 27. The average variation in the count of hairs was 18, with a range of from 2 to 12. It is evident that the variation between analysts in making the count of bran particles is greater than in making the count on hairs. DAILY VARIATION DUE TO CONDITION OF LIGHT. ETC. To determine what influence, if any, physical conditions, such as degree of light, have upon the count, it is necessary to first eliminate, as far as possible, the personal variations already considered. This may be accomplished by taking the average of two counts on three ' By " daily average" is meant the average of two counts made by the same analyst on the same slide on a given day. 8 BULLETIN 839, U. S. DEPARTMENT OF AGRICULTURE. slides for the same day and averaging the three results to determine the analyst's daily variation. This is calculated for each analyst. The ultimate daily variation is the average of the daily variation of the two analysts computed for each day. The daily variation for each analyst is shown in Table 4. Table 4. — Daily variation for each analyst. Variation. Bran particles. Hairs. SUde. Jan. 7, 1918. Jan. 8, 1918. Jan. 9, 1918. Jan. 7, 1918. Jan. 8, 1918. Jan. 9, 1918. A Keenan. 89 60 84 77 108 83 88 93 85 79 54 63 65 98 81 86 88 76 81 51 68 66 105 77 86 86 76 61 59 68 62 55 61 59 58 60 67 51 67 61 55 53 62 56 63 65 B .. 59 c 68 64 A . Lyons. 54 B 55 C 63 Analyst's daily 57 U Itimate daily 60 The results in Table 4 seem to indicate that on January 7, 1918, there was a tendency to count higher on bran particles than on the other days. It is believed, however, that this was in whole or in part due to the clearing action of the glycerin employed to preserve the slides for counting on subsequent days, which tended to make the identification of the bran particles more difficult after the first da}'. SLIDE VARIATION DUE TO LIMITS OF ACCURATE WEIGHING OF THE TEST PORTION OF FLOUR. In order to determine the absolute variation between the slides, it is evident that an average must be obtained from which the personal variations and the daily variations have been eliminated as far as possible. This is accomplished by computing for each slide the aver- age of all counts made on bran particles, and also making a similar computation for the hair count (Table 5). Table 5. — Counts ofhran particles and hairs on slides. Bran particles. Hairs. Slide A. Slide B. Slide C. Slide A. Slide B. Slide C. 89 79 81 108 98 105 60 64 51 83 81 77 84 63 68 88 86 86 61 67 65 55 55 54 59 61 59 61 53 55 68 67 68 59 62 63 193 167 1 79 159 156 164 » Average slide count. MICROSCOPICAL EXAMINATION OF FLOUR. 9 The variation in the counts on these slides naturally raises the ques- tion of the limits of accuracy in weighing out the test portion of flour. vSince the amount of flour used on a slide is 5 milligrams, it is desirable to determine how great is the error due to weighing the test portion of flour. The balance employed in this investigation was a fine assay balance. In weighing the sample the vibration method was used, and the quantity of flour was so adjusted as to produce a deviation of approximately not more than one-fourth of a space on each side of the zero point of the scale. This is equivalent to not more than 1/40 milligram, or one-half of 1 per cent, on the basis of the portion of flour used (5 milligrams) . Hence any error in weighing can not be accepted as an explanation of the difference in slide counts. VARIATION IN HOMOGENEITY OF BULK SAMPLE. The question has been raised as to whether or not a portion of the slide variation might not be accredited to lack of uniformity of the bulk sample, due to the fact that any grade of flour is usually the component result of several constituent streams which vary more or less among themselves. The fact that in general practice the flour stocks are subjected to a certain degree of purification, however, leaves this factor little chance to figure to any great extent. This point was tested by passing a certain sample of flour which had an average count of 32 bran particles and 64 hairs through a 30-mesh sieve and making up and counting 12 slides. The bulk sample was then passed through the sieve once more (making two times for the sample), and another series of slides made and counted. Finally, the sample was put through the sieve twice more (making four times for the sample) , and a third series of 12 slides made and counted. The results of these tests are given in Table 6, the counts in which are the average of the results obtained by two persons. Table 6. — Effect of variation in homogeneity of sample on count. Sample passed through 30-mesh sieve- Once. Twice. Four times. Bran particles. Hairs. Bran particles. Hairs. Bran particles. Hairs. 31 48 36 33 41 36 35 35 37 35 30 34 73 73 58 53 64 64 74 75 57 76 66 60 26 37 25 22 27 32 32 27 34 26 39 30 67 70 58 54 61 57 83 64 61 72 67 67 21 22 30 32 23 29 36 34 39 38 33 37 76 53 56 . 81 79 66 60 56 70 52 65 48 135 2 18 166 2 23 '29 217 165 2 29 131 2 18 163 2 31 152332°— 20- 1 Average. -Bull. 839- > Variation 10 BULLETIN 839, U. S. DEPARTMENT OF AGRICULTURE. Apparently, sifting or thorough mixing of the flour a number of times has little appreciable effect upon the offal count obtained. NUMBER OF SLIDES COUNTED. In practice, two slides, or at most three, from the sample of flour have been used as the basis for judgment as to the character of the product as far as the offal material was concerned, and the question might very properly be asked if that number is sufficient. In order to test out this point, 12 slides were prepared from the same bulk sample of flour. Two counts on each slide were made of the bran particles and hairs by each of two analysts. The results obtained are recorded in Table 7. Table 7. — Counts on 12 slides. Slide des- ignation. Analyst. Count No. Bran parti- cles. Hairs. Slide des- ignation. Analyst. Count No. Bran parti- cles. Hairs. A... Keenan ....do Lyons . ..do 1 2 1 2 1 2 1 2 1 2 1 2 1 2 1 2 1 2 1 2 1 2 1 2 21 22 20 22 24 23 19 23 23 31 34 33 33 30 31 35 24 22 23 26 31 29 29 30 76 75 78 77 54 53 52 56 55 60 59 53 77 81 84 82 79 79 82 78 64 65 70 66 G Keenan do 1 2 1 2 1 2 1 2 1 2 1 2 1 2 1 2 1 2 1 2 1 2 1 2 38 32 33 40 37 24 38 40 26 44 44 42 39 38 34 41 32 30 33 37 32 36 42 39 64 A G 57 A... G Lyons do 61 A G 61 B . Keenan .do H Keenan do 61 B H 47 B Lyons .do H Lyons do 56 B... H 61 C Keenan do Keenan do 74 c 67 C Lyons do 1 1 Lyons I ' do 67 C 72 D Keenan ... do Keenan 50 D do 55 D Lyons do Lyons do 48 D 56 E... Keenan do K Keenan do 63 E... . K 60 Ei Lyons do K Lyons do 68 E... K 71 F Keenan .do L Keenan do 47 F L 49 F.. . . Lyons do L Lyons L ' do 47 F 51 From the data in Table 7 it is possible to average Keenan's first count on slide A with each count made by him on each of the other slides. By averaging the slides by two, 20 is found to bo the lowest average and 43 the highest average for bran particles, considering Keenan's results only. If the average of counts for three slides instead of two is to be taken as the basis for final judgment of the product, it is apparent that 22 is the average of the three lowest results and 40 the average of the three highest (Keenan's results on bran particles). Taking the average of the counts on each of four slides gives an average minimum count of 22 and an average maximum count of 39. Table 8, based on data obtained from Table 7, has been prepared to show the results of such methods of grouping. MICROSCOPICAL EXAMINATION' ..i" FI.OUR. Table 8. — Effect of method of com,'pvi'>"rvM KS. Before undertaking a discussion of th*' vv,.rk .Ittue on finished com- mercial flours, it seemed advisable to consI'l'T ti.f; degree of purity of the various mill stocks entering into li..- ■ ..Mi,.-.sition of the end- product. The data compiled in Table '' ■i-in-.n-irate the quality of the stocks made on the break rolls, the i>ttri....:, of which is to crush the wheat kernel to release the enclosed «(!'Si.r:iQ that is later re- duced to fineness on other rolls and firi.t!lv inniiied of offal debris. The general practice in milling is to make .iv lii 1 1,> break flour as possi- ble. When break flour is made to an\ >. tains a notable amount of offal, consist "nv as numerous hairs from the beard. Thf- r-' were obtained on samples of material pror iM-an particles, as weU n]\< recorded in Table 9 • I I'f'vm the first, second, Silk number. i j .W^sl,.:.' i lOxx !..:• llxx ,; 12xx . 1 ■• 12x 12XXX 13xx i l->r. 14xxx. ' The designations for the various stocks and grades of flour 6 •..;'.;., ,^1. .> /'ell as the statements concern- ing the kind of wheat from which the Hour was milled, weret».l.-en irom t.h« miilers supplying the saniplH.. and were not verified in the Bureau of Chemistry. 12 BULLETIN 839, U. S. DEPARTMENT OF AGRICULTURE. Table 9. — Results of examination of -products from, break machines. Sample No. Type of wheat. Bolting cloth. Bran particles. Hairs. 196 165 186 83 117 43 334 162 76 61 46 64 113 38 182 58 324 42 166 65 216 162 105 23 150 83 47 44 32 54 142 58 75 38 138 31 120 121 628 107 367 144 159 26 118 53 68 60 37 56 375 73 131 53 135 46 810 213 322 116 262 57 118 50 132 147 228 106 285 66 Total. 11079-K-A. 17146-L-B. 17125-L-C. 17128-L-A. 17159-L-A. 17173-L-A. 17165-L-O. 17167- L-A. 17143-L-B. 17146-L-A. 17190- L-B. 17125-L-D. 17128-L-C. 17159- L-B. 17173-L-B. 17133- L-F . 17165-L-P. 17167-L-B. 11079-K-D. 17143-L-C.. 17146-L-D. 17125-L-E. 17128-L-D.- 17159-L-C.. 17173-L-C.. 17133-L-G. 17165-L-Q. 17167-L-C.. 17143-L-D. 17146-L-C. 17125-L-F. 17128- L-E. 17173-L-D. 17165-L-R. 17167-L-D. Hard do Hard and soft. do do do Soft do Hard do do Hard and soft. do do do Soft do do Hard do do Hard and soft. do do do Soft do do Hard do Hard and soft. do do Soft do FreST BREAK MATEBIAL. (?) 113CX, 13xx (?) (?) (?) (?) lOxx 12xxx, 14xxx SECOND BREAK MATERIAL. lOx, llx 12xx, 13xx 12xx, 13xx (?) (?) ■-• (?) (?) (?) lOxx 12xxx, 14xxx THIRD BREAK MATERIAL. (?) lOx, llx 12xx, 13xx (?) (?) (?) (?) (?) lOxx 12XXX, 14xxx FOURTH BREAK MATERIAL, llx, 12x 12xx, 13xx (?) (?) (?) lOxx 14xxx 361 269 160 496 137 110 151 240 366 231 378 128 233 91 86 200 113 169 241 735 511 185 171 128 93 448 184 181 1,023 438 319 168 279 334 351 For the purpose of comparison, the data from Table 9 have been summarized in Table 10. Table 10. — Summary of results of examination of products from break jnachines. Machine stock. Average. Bran particles. Hairs. First break 156 139 213 308 84 60 73 122 Third break The offal content of the break roll products is high, as would be expected. A microscopical examination is hardly necessary to establish this fact. The fluffy and dirty appearance of such products, even from casual examination, is sufficient to show that they are of low quality, judging from the offal material present. MICROSCOPICAL EXAMINATION OF FLOUR. 13 Tests similar to those made on break roll products were made on middlings stock. Middlings are usually recognized as being the medium granular particles of the endosperm resulting from the crack- ing of the wheat kernel on the break rolls. After proper purification or removal of the branny material, the middlings are milled, on the reduction rolls, to the fineness of flour. The results of experimental work done on middlings stocks are recorded in Table 11. Table 11. — Results of examination of middlings stocks. Sample No. 15196- K-E. 17144-L-I.. 17190-L-I.. 17125-L-L. 17159-Iv-F . 17132-L-J.. 17133-L-L. 17165- L-B. 17185-L-C . 15196- K-C. 17144-L-J.. 17146-L-O. 17190-L-I.. 17125-L-N. 17159-L-G. 17132-L-K. 17133-L-O. 17165- L-C . 17185- L-D. 17144- L-K. 17146-L-I.. 17190-L-J.. 17125-L-P. 17128-L-K. 1715»-L-H. 17173-L-O. 17132-L-L. 17133-L-Q. 17165- L-D. 17167-L-J.. 17185- L-F. 15196- K-J. 17190-L-K. 17125- L^T. 17128-L-O. 17159- L-I.. 17171-L-L. 17132-L-M. 17133-L-S. 17165-L-F . 17167-L-M. 17167-L-K. 15196- K-K. 17144- L-M., 17146-L^H.. 17190-L-L.. 17128- L-P.. 17159-I^J... 17171-L-M.. 17173-L-P., 17133-Lr-U.. 17165- U-H. 17167- L-O.. Type of wheat. Hard do do H.ird and soft. do Soft do do do Hard do ....do do Hard and soft. do Soft ....do ....do ....do Hard ....do ....do Hard and soft. ....do ....do ....do Soft do do do do Hard do Hard and soft. do do do Soft do do do do Hard do ....do do Hard and soft. ...do ....do do Soft do ....do Bolting cloth. rmST MIDDLIKGS STOCK. lOxx, llxx, 12xx lOx lOxx (?) (?) (?) (?) ; lOxx lOxx SECOND MIDDLINGS STOCK 12xx, 13xx, 14xx lOx llxx, 12xx llxx (?) (?) (?) (?) lOxx lOxx THIRD MIDDLINGS STOCK. lOx, llx llx, 12xx lOxx, llxx (?) (?) (?) (?) (?) (?) lOxx 14xxx lOxx FOURTH MIDDLINGS STOCK llxx, 12xx, 14xx llxx, 12xx (?) (?) (?) llxxx, 12xx (?) (?) 12xx (?) (?) FIFTH MIDDLINGS STOCK. llxx, 12xx, 14xx lOx, llx llxx, 12xx llxx, 12xx (?) (?) 12xx (?) (?) 12xx (?) Bran particles. 5 7 29 100 11 7 25 59 48 10 76 7 38 8 82 26 36 29 115 40 Hairs. Total. 10 31 150 12 12 26 65 75 12 10 101 8 61 18 01 27 31 31 139 4« 31 10 20 9d 110 10 84 114 03 «T 53 14 BULLETIN 839, U. S. DEPARTMENT OF AGRICULTURE. Table 11. — Results of examination of middlings stocks — Continued. Sample No. 15196- K-G... 17144-L-N... 17190- L-M... 17125- L-BB. 17128- L-Q... 17173- L-Q... 17173-I^R... 17133- L-W.. 17165-L-I.... 15196-K-F. 17144- L-0. 17190-L-N. 17128- L-R. 17159- L-M. 17133- L-X. 17165- I^J.. 17167- L-L. 17190- L-0. 17173- L-S.. 17165-L-K. 17167-L-S. Type of wheat. Hard do ....do Hard and soft. do do do Soft do Hard do do Hard and soft. do Soft do do Hard Hard and soft. Soft Soft. Bolting cloth. SIXTH MIDDLINGS STOCK. llxx, 12xx, 13xx, 14xx llx, 12x, 13xx llxx, 12xx (?) (?) (?) (?) (?) 12xx SEVENTH MIDDLINGS STOCK 12xx, 13xx, 14xx 12x, 13x 13xx, 14xx (?) (?) (?) 12xx, 13xx (?) EIGHTH MIDDLINGS STOCK. 13xx, 14xx (?) 13xx, 14xx NINTH MIDDLINGS STOCK. (?) Bran particles, 26 24 139 87 70 41 22 140 60 16 36 119 63 104 194 143 45 137 51 264 Hairs. 25 Total. 56 25 172 102 107 107 57 158 25 40 162 89 182 210 1G« 58 160 103 302 117 The average results obtained on the middhngs stocks examined have been summarized in Table 12. Table 12. — Summary of results of examination of middlings stocks. Stock. First middlings.-.. Second middlings- Third middlings... Fourth middlings. Fifth middlings... Sixth middlings... Sevonth middlings Eighth middlings.. Ninth middlings... Average. Bran particles. 24 29 21 41 46 65 90 150 92 Hairs. Total. 33 39 30 61 64 91 116 187 117 The results in Table 12 clearly demonstrate that the middhngs stocks are much cleaner than stocks obtained from the break rolls. The first five middlings stocks average low in the total offal count, while the stocks from the sixth to ninth middlings, inclusive, average appreciably higher. In other words, the more thorough the purifi- cation process, the lower will be the offal count. For the purpose of showing the offal count on the stocks which pass into some so-called patent flours, three different sets of mill streams MICROSCOPICAL EXAMINATION OF FLOUR. 15 were examined, these streams being designated as entering into the composition of certain finished flours. The mill streams composing such flours were milled from hard, blended, and soft wheats, respec- tively. The results of these examinations are shown in Tables 13, 14, and 15. Table 13. — Resttlts of examination of mill streams composing a patentflour {sample No. 171U-L-FF) milled from hard wheat. stock. Bran particles. Hairs. 14 4 7 3 8 1 19 2 9 1 24 1 36 4 36 5 30 5 59 8 37 2 151 20 13 2 Total. First middlings Second middlings Third middlings Fourth middlings Fifth middlings Sixth middlings Seventh middlings Middlings Do First sizings Second sizings Sizings Finished flour (70 per cent patent)' IS 10 9 21 10 25 40 41 35 67 39 171 IS > This finished flour Is composed of the stocks described above it. Table 14. -Results of examination of mill streams composing a patentflour (sample No. 17159-L--V) milled from blended wheat. stock. Bran particles. Hairs. Total. First break Second break Third break Break chops Do First middlings Second middlings Third middlings Fourth middlings Fifth middlings (head) Fifth middlings (tail) Coarse tailings Coarse sizings Finished flour (70 per cent patent) » 137 91 128 95 140 40 12 41 18 19 33 32 14 35 I This finished flour is composed of the stocks described above it. Table 15. — Results of examination of mill streams composing a patentflour (sample No. ni32-L-U) m,illedfrom soft wheat. First middlings Second middlings 25 1 26 Third middlings 9 9 Fourth middlings 26 1 27 Fine sizings 10 1 11 Medium sizings 21 2 23 Coarse sizings 14 2 16 Finished flour (60 per cent patent)' 19 1 20 1 This finished floor is composed of the stocks described above it. Bran particles. Hairs. 19 3 25 1 9 26 1 10 1 21 2 14 2 19 1 Total. <1'6 BULLETIN" 839, U. S. DEPARTMENT OF AGRICULTURE. It is interesting to observe the variety of streams drawn upon for the composition of different so-called patents, as well as the varia- tion in the offal count of the stocks employed in the milling of such finished flours. If space permitted, additional information could be submitted to illustrate how variable the different mill stocks are as far as offal content is concerned. In many instances where lower-grade stocks have been employed in making a flour, however, the finished product has usually been purified sufficiently to cause the resultant offal count to be appreciably low. And in many cases the contrary is true. EXAMINATION OF COMMERCIAL GRADES OF FLOUR. The assembled flours employed in this part of the investigation were collected by B. C. Winslow, food and drug inspector. Bureau of Chemistry, United States Department of Agriculture. As these flours were milled under a variety of conditions, they necessarily reflect such conditions in the finished product. The inspector gave the following statement as to the designations applied to these flours: "As a general thing, these names were used in harmony with the usage of the mill where they were taken. The method of assembling, with the streams, percentages, etc., were given when feasible, and as correctly as possible from the information available. The general terms 'patent,' 'clear,' and 'straight' were used to clas- sify in a general way the assembled grades of flour, and vary mth each mill." With this information in mind, an attempt was made to apply the microscopical method already described to an examination of these products for the purpose of developmg a system for the classification of flours based on the offal content. A detailed discussion of the actual data obtained from these tests, with a general summary on the various so-called grades, follows. PATENT FXOURS. PATENT FLOURS MILLED FROM HARD WHEATS. Thirty-six patent flours said to have been milled from hard wheats were examined microscopically, and theij bran particle and hair count determined. The commercial grade designations ranged from 40 to 94 per cent. In some instances the flour had been bleached; in others it was bleached only lightly or not at all. Table 16 gives the results of this examination. MICROSCOPICAL. EXAMINATION OF FLOVR. 17 Table 16. — Results of examination of patent flours milledfrom hard wheats. Sample No. Commer- cial grade. Bleacheu. Bran particles. Hairs. Total. 15163-K-R "Per cent patent." (?) (?) 40 52 58J 60 65 68 68 70 70 70 71 72 72 74 74 75 75 75 75 75 77 78 79 80 80 80 83 83 83 84 85 85 88 94 Yes 16 29 19 72 27 15 24 23 22 22 16 17 20 17 27 29 13 66 33 33 30 19 25 35 25 54 44 17 34 34 36 33 32 36 33 62 8 13 13 45 9 5 20 20 15 21 12 11 10 4 14 26 2 33 2 36 39 13 24 28 31 28 9 19 16 10 30 12 23 15 9 34 24 15178-K-U (?) 42 17151-L-MM No 32 17151-L-NN No 117 17150-L-T No 36 15152-K-A Yes 20 15134-K-EE No 44 11078-K-LL No 43 11078-K-MM Yes 37 11070-K-rF (?) 43 1.5112-K No 28 15113-K Lightlv 28 17154-L-AA No 30 15174-K-LL (?) 21 15187-K-X No 41 15170-K-CC No 55 17144-L-FF No 15 1719(>-L-CC Yes 99 17143-L-CC No 35 15193-K-EE No 69 15193-K:-DD Yes 69 11028-K-B (?) 32 17157-L-A Yes . . . 49 17183-L-A No 63 17184-L-Q Yes 56 11064-K-A No 82 17175-L-MM No 53 No 36 17147-L-BB Yes 50 17148-L-MM Yes 44 17156-L-FF Yes 66 17155-L-JJ . No 45 55135-K:-CC No 55 17145-L-B No 51 17111-L-S No 42 17180-L-JJ . Yes 96 On these hard-wheat patents the bran particle count ranged from 15 to 72, with an average of 30. The hair count ranged from 2 to 45, with an average of 18. The total offal count ranged from 15 to 117, with an average of 45. PATENT FLOURS MILLED FROM SOFT WHEATS. The patent flours milled from soft wheats are more starchy than those milled from hard wheats. This starchy character is manifest even when the sample of flour is poured out upon a piece of paper. The soft-wheat flour will not ''flow" like a flour made from hard wheat, but is more "powdery" and starchlike rather than granular, as in the case of hard-wheat flours. Thirteen patent flours stated to have been milled from soft wheats were examined microscopically. As in the case of hard-wheat flours, the commercial grades, as indi- cated by percentages, varied markedly, and can be regarded only as approximate. The percentages ranged from 35 to 90 per cent. Some of the flours were bleached, others lightly bleached, and still others not bleached at all. Table 17 gives the results of this examina- tion. 18 BULLETIN 839, U. S. DEPARTMENT OF AGRICULTURE. Table 17. — Results of examination of patent flours milled from soft wheats. Sample No. Commer- cial grade. Bleached. Bran particles. Hairs. TotaL 17161-L-A "Per cent patent." (?) 35 40 45 60 60 60 65 65 65 67 75-80 90 Yes 72 32 32 33 19 49 23 50 56 133 46 53 51 10 25 11 32 1 34 17 12 22 29 19 26 30 82 17189 L No 5T 17167 L FF Yes 43 17165-L EE . . Yes 65 17132 L-U No 2* 17169 L S (?) 83 17187- L-V No 40 15121 K EE Yes 63 15126 K-FFF Yes 78 17133 L FF Yes 163 17161-L-LLL No 65 17164-L-T. (?) 7» 11007-K (?) 81 The bran particle count varied from 19 to 133, the hair count from 1 to 34, and the total offal count from 20 to 162. The average count for bran particles was 49 and that for hairs 20, while the average total offal count amounted to 70. PATENT FLOURS MILLED FROM BLENDED WHEATS. The flours classified under blends were manufactured from mix- tures of hard and soft wheats. Similar information was obtained for these flours as for the hard and soft types. The designations for the so-called grades varied from 70 to 85 per cent. Of the 12 sam- ples examined, 4 were bleached and 8 unbleached. Table 18 gives the results. Table 18. — Results of examination of patent flours milled from blended wheats. Sample No. Commer- cial grade. Bleached. Bran particles. Hairs. Total. 11084-K "Percent patent." 70 70 70 70 70 75 75 80 82 83 85 85 No 29 32 31 20 51 40 18 36 63 61 47 83 13 13 18 15 25 37 13 19 27 40 21 17 43 11085-K Lightly 45 11086 K . . Heavily 4» 17159- L-V No ;« 17171-L-B No 78 17168- L-YS . . ... . No 77 17168- L-YYH No 31 17179-I>-YY No No Yes Yes No 55 17127-L-S 90 17116-L-D 101 17123-L-FF 68 17125-L-FF 100 The bran particle count ranged from 18 to 83, with an average of 42. The hair count ranged from 13 to 40, with an average of 21. The total offal count ranged from 31 to 101, with an average of 64. MICEOSCOPICAL, EXAMINATION OF FLOUR. 19 PATENT FLOURS MILLED FROM MIDDLINGS STOCKS ONLY. Information was obtained concerning the history of the mill streams entering into the composition of a large number of so-caJled patent flours. The data collected showed that middlings stocks only were employed in composing these flours. The results of the counts made on these samples are recorded in Table 19. Table 19. — Results of examination of patent flours viilledfrom middlings stocks only. Sample No. Commer- cial grade. Bleached. Bran particles. Bairs. Total. 17151-L-MM 15152- K-A.. 11070- K-PF. 17154-L-AA . 17158- L-X... 15186-K-X.. 15174-K-LL. 1S170-K-CC. 17144-L-FF. 15181-K-S.. 15146- K-W. 15163-K-R.. "Per cent patent." 40 60 70 71 71 72 72 74 74 80 Short patent. (?) Yes (?). No. Yes No. (?). No. No. Yes (?). Yes Table 19 shows that the bran particle count ranged from 13 to 29, with an average of 20, that the hair count ranged from 2 to 26, with an average of 13, and that the total offal count ranged from 15 to 55, with an average of 33. These results demonstrate the fact that the purified middlings stocks employed had some effect upon the purity of the end-product. From the information the writers were able to obtain, however, so-called patent flours were not always composed of the best streams in the mill. PATENT FLOURS MILLED FROM MIDDLINGS STOCKS PLUS LOWER-GRADE STOCKS IN THE MILL. As already stated, stocks other than first-class middlings were often passed into patent flours. According to the information submitted, break flours and lower grades of middlings frequently were found to have been employed in the manufacture of the finished flour. The results recorded in Table 20 illustrate the effect of the addition of mill streams appreciably high in offal to the finished product. 20 BULLETIN 839, U. S. DEPARTMENT OF AGRICULTURE. Table 20. — Results of examination of patent flours milled from, middlings stocks in addition to lower-grade stocks in the mill. Sample No. Com- mercial grade. Variety of wheat. Bleached. Bran particles. Hairs. Total. 17189-L-O 'Percent." 35 52 60 65 67 70 70 75 75 75 75 75-80 78 79 80 80 82 83 83 83 84 85 85 85 94 Soft No 32 72 49 133 46 20 51 30 33 66 18 S3 35 25 54 36 63 34 36 61 33 47 83 32 62 25 45 34 29 19 15 25 39 36 33 13 26 28 31 28 19 27 16 30 40 12 21 17 23 34 57 17151-L NN Hard No 117 17169-L-S Soft Yes 83 17133- L-FF do Yes 162 17161-L-LLL .. . do No 65 17159- L-V Blend. No 35 17171-L-B do No 76 15193-K-DD Hard Yes 69 15193-K-EE do . . do No 69 17190-L-CC Yes 99 17168-L-YYH... Blend No 31 17164-L-T Soft Yes 79 17183-L-A Hard No 63 17184-L-Q do No 56 11064-K-A do No 82 17179-L-YY Blend No 55 17127-L-S .. . do Yes 90 17147-L-BB Hard Ye.s 50 17156-L-FF do Yes 66 17116-L-D Blend Yes 101 17155-L-,TJ Hard No 45 17123-L-FF Blend Yes 63 17125-L-FF . . do No 100 15135-K-CC Hard No 55 17180- L-JJ. .. do Yes 96 The total offal count on these samples was consistently higher in most cases than the results obtained on samples ground from mid- dlings stock only. The addition of break flour stocks appeared to have a marked effect upon their quality with respect to the offal count. The bran particles ranged in count from 18 to 133, with an average of 48. The hair count ranged from 12 to 45, with an average of 26. The total offal count varied from 31 to 162, with an average of 74. GENERAL CONCLUSIONS ON PATENT FLOURS. 1. The commercial grades of so-called patent flours ranged from 35 to 90 per cent. These percentage figures apparently were in- tended to indicate that a certain percentage of the total flour content of the wheat kernel passed into this grade, the remainder being employed in other grades. 2. The average total offal count obtained on all commercial patent flours examined was 57. 3. Patent flours showed a marked variation in the total offal comit obtained on different samples from various mills. 4. The limitations and the average counts on bran particles and hairs have been briefly summarized in Table 21. MICROSCOPICAL EXAMINATION OF FLOUR. 21 Table 21. — Limitations and average counts on bran particles and hairs for patent flours . Commercial grade. Bran particles. Variation. Average Hairs. Variation. Hard-wheat patent Soft- wheat patent Blended-wheat patent 13 to 72. 19 to 133 18 to 83. 2 to 45., 1 to34., 13 to 40 STRAIGHT FLOURS. When only one grade of flour is manufactured in the mill, this grade is commercially designated as a straight flour, if it contains the entire flour content of the wheat that it is possible to mill. It might be considered to contain all of the flour that could be obtained from the wheat kernel with the exception of a certain percentage of so- called low-grade or red dog flour. Such a straight flour naturally would contain more of the branny particles from the wheat kernel than would a patent flour. The practice of compositing such a flour apparently varies in different mills. Tests were made upon a large number of straight flours milled from hard, soft, and blended wheats. The detailed information on these tests is given in the following para- graphs. STRAIGHT FLOURS MILLED FROM HARD WHEATS. Seventeen straight flours reported as having been milled from hard wheats were examined for their offal content. The com- mercial grades ranged from 92 to 100 per cent. The results of the examination appear in Table 22. Table 22. — Results of examination of straight flours milled from hard wheats. Sample No. Com- mercial grade. Bleached. Bran particles. Hairs. Total. 15196-K-U "Per cent straight." 92 95 95 96 97 97i 98 98 98 98 98 98 98 98 100 100 100 No 33 71 50 89 37 57 62 55 58 62 71 63 57 71 76 60 121 34 55 45 33 25 39 31 51 61 87 65 19 26 47 61 17 22 67 11028-K-E (?) 126 17157- L-B Yes 95 17155-L-HH Yes 122 15154-K-C . Yes 62 15106-K (?) 96 11067-K (?) 93 15136-K-BB Yes 106 15147-K Yes 119 15191-K (?) 149 15194-K-U Ves 136 17113-L Yes 82 17152-L-Y (?) 83 17177-L-XX . No 118 11073-K-GG (7) 137 17146-L-F No 77 17186-L-F No 143 The count obtained on bran particles ranged from 33 to 121 and that on hairs from 17 to 87. The average bran particle count was 64 and the average hair count 43. The total offal count ranged from 62 to 149, with an average of 106. 22 BULLETIN 839, U. S. DEPARTMENT OF AGRTCITLTURE- STRAIGHT FLOURS MILLED FROM SOFT WHEAT.S. Seventeen straight flours reported to have been milled from soft wheats were examined. The commercial grades ranged from 90 to 100 per cent. Table 23 gives the results of this examination. Table 23. — Results of examination of straight flours milled from soft wheats. Sample No. Commer- cial grade. Bleached. Bran parti- cles. Hairs. Toliil. 11096-K "Per cent straight." 90 90 90 90 90 90 97 100 100 100 100 100 100 100 100 No 52 41 56 92 89 50 111 119 55 109 153 93 97 109 52 34 92 ;o 31 38 58 26 60 70 54 27 71 81 40 22 34 39 34 38 92 11097-K... Light! v.. . 72 11098-K 94 15125-K-BB 150 15126-K-DJ>D . . Yes 116 17166- L-Q Yes 110 15125-K-J"F .... 181 15125- K-J.I No 173 17188-L-X No 82 15125- K-Y Yes 180 15125-K-OO .. Yes.. 234 15126-K-A\A .. . . Yes 133 17136-L-Z Yes 119 17165- L-A A Yes 143 17176-L-W No 91 17185-L-H No 68 17186-L-C... . No 130 The bran particle count varied from 34 to 153, with an average of 82, and the hair count varied from 22 to 81, with an average of 45. The total offal (;ount ranged from 68 to 234, with an average of 127. STBAIGHT FLOURS MILLED FROM BLENDED WHEATS. Eighteen samples of flour stated to have been milletl from blends of hard and soft wheats were examined for their offal content, as in the case of the hard and soft types. The commercial grade desig- nations varied from 90 to 1 00 per cent. The results of the exam- ination are given in Table 24. Table 24. — Results of exam.ination of straight flours milled from blended wheats. Sample No. Commer- cial grade. Bleached. Bran parti- cles. Huirs. Total. 11087-K "Per cent straight." 90 90 90 90 90 90 90 90 96 97 97i 97i 97* 97J 96i 98 100 (?) No 50 51 50 183 21 52 41 56 90 98 42 43 52 73 83 33 88 86 26 22 28 18 36 40 31 38 47 30 28 29 26 37 45 47 58 37 76 11088-K Lightly 73 11089-K Heavily 78 17118-L-J No 201 17173-L-V Yes 57 11096-K No 93 11097-K.... Lightly 72 11098-K Heavily 94 17120-L-N Yes m 17121-L-SS 11090-K.... No No . 128 70 11091-K Lightly 72 11092-K Heavily Yes No No No No 78 17117-Iv-DD 110 17115-L-E 128 17173-L-W 80 15195-K-A 17128-T,-Z... 146 123 MICROSCOPICAL EXAMINATION OF FLOUK. 23 The bran particle count varied from 33 to 183, with an average of 68, while the hair count varied from 18 to 58, with an average of 34, The total offal count varied from 57 to 201, with an average of 100. The average total offal count obtained for the straight flours was 111, as against 57 for patent flours. MILL STREAMS EMPLOYED IN THE MANUFACTURE OF CERTAIN STRAIGHT FLOURS. Data were obtained on the mill streams employed in the manu- facture of certain straight flours, and these streams were examined for their offal content for the purpose of illustrating the quality of the material sometimes used in making up such flours. The results are given in Tables 25, 26, and 27. Table 25. — Results of examination of mill streams employed in the manufacture of a straight flour {sample No. 17146-L-F) milled from hard wheats. Stock. Bran particles, Hairs. Total. First break Second break Third break Fourth break Fifth break Second middlings Third middlings Third middlings (second stream) Fifth middlings Cut-off flour Cut-off flour Chimk flour Second chunk flour TaiUngs flour Tailings flour 100 per cent straight flour' 1S6 166 367 322 456 29 27 13 21 15 76 308 50 76 155 60 83 65 144 116 176 2 6 4 5 4 18 90 5 24 47 17 269 231 511 438 632 31 33 17 26 19 94 398 55 100 202 77 ' Composited from the mill streams listed above it. Table 26. — Results of examination of mill streams employed in the manufacture of a straight flour (sample No. 17165-L-AA) milled from soft wheats. Stock. Bran particles. Hairs. Total. First break Second break Third break First, second, and third breaks Fourth break Fifth break First middlings Second middlings Third middlings Fourth middlings Fifth middlings Sixth middlings Seventh middlings Eighth middlings First germ flour 100 per cent straight flour ' 113 75 131 101 228 368 21 48 26 29 55 60 143 264 50 109 38 38 53 45 106 173 8 27 7 2 12 18 23 38 5 34 151 113 184 146 334 641 29 75 33 31 67 78 166 302 55 143 ' Composited from the mill streams listed above it. 24 BULLETIN 839, U. S. DEPARTMENT OF AGRICULTURE. Table 27. — Results of examination of mill streams employed in the manufacture of a straight flour {sample No. 17128-L-Z) milled from blended wheats . stock. First break Second break Third break Fourth break Fifth break First middlings Second middlings.. Third middlings. . . Fourth middlings. . Fifth middlings Sixth middlings . . . Seventh middlings. First sizings Second sizings First tailings Second tailings Head cuts Tailcuts Straight flour 1 Bran particles. Hairs. 334 150 118 118 296 66 41 69 38 74 70 63 56 107 134 108 132 130 87 162 83 53 50 101 31 21 26 23 36 37 26 11 34 43 48 63 70 37 Total. 496 233 171 168 397 97 62 95 61 110 107 89 67 141 177 156 195 200 124 1 Composited from the mill streams listed above it. GENERAL CONCLUSIONS ON STRAIGHT FLOUES. 1. The commercial grades of so-called straight flours ranged from 90 to 100 per cent. 2. The average total offal count obtained on all commercial straight flours examined was 111. 3. Straight flours showed a decided variation in the total offal count obtained on different samples from various mills. CLEAR FLOURS, Clear flour, so-called, is often considered among millers as being a mixture of odds and ends of the milling stocks. Low grades of mid- dlings and break flours often pass into it, although frequently it con- tains the purest quality of middlings stock from the tail of the mill. Clear flours which were said to have been milled from hard, soft, and blended wheats, respectively, were examined. CLEAR FLOURS MILLED FROM HARD WHEATS. Thirty-one clear flours stated to have been milled from hard wheats were examined. Their percentages ranged from 6 to 52. Table 28 shows the counts thus obtained. MICROSCOPICAL EXAMINATION OF FLOUR. 25 Table 28. — Results of examination of dear flours milled from hard wheats. Sample No. Bleached. Bran particles. Hairs. 331 132 238 166 306 50 191 98 197 77 156 126 294 223 181 102 271 184 241 62 193 136 127 119 65 39 82 68 71 67 131 124 410 196 172 140 193 204 158 102 316 71 271 93 92 71 79 57 77 49 127 178 268 43 118 133 126 114 151 147 72 45 Total. 17180-L-KK. 17151-L-OO. 17142-L-EE. 17112- L-T... 17150-L-U. . . 15138-K-DD. 17154-L-CC.. 17145-L-A... 17147-L-AA . 17175-L-NN. 17183-L-B... 17184-L-P... 11065-K-A... 11079-K-JJ.. 11079-K-KK 15169-K-DD. 15192-K-FF . 15186- K-Y . . n028-K-C... 15175-K-MM 17143-L-BB. 17144- L-II... 15115-K 15n6-K 15117-K 11071-K-EE. 17186-L-E... 15150-K-AA. 15137- K-FF. 15180- K-A A. 17151-L-NN. 'Per cent clear." 6 8 10 12 12 13 14 15 15 15 16 16 18 22 22 23 23 24 25 25 25 26 27i 27i 27J 30 30 35 33-35 18 52 Yes No No No No No No No No No No No (?) No Yes (?) Yes.... No (?) (?) No No No Lightly- Heavily (?) No Yes.... (?) (?) No 463 404 356 289 274 282 517 283 455 303 329 246 104 150 138 255 606 312 397 260 387 364 163 136 126 305 311 251 240 298 117 The bran particle count on these samples varied from 65 to 331, with an average of 174. The hair count ranged from 43 to 223, with an average of 109. The total offal count varied from 104 to 517, with an average of 295. CLEAR FLOURS MILLED FROM SOFT WHEATS. Thirteen samples of clear flour reported to have been milled from soft wheats were examined, these samples varying from 5^ to 50 per cent as far as commercial grades are concerned. Table 29 gives the results obtained. Table 29. — Results of examination of clear flours milled from soft wheats. Sample No. Com- mercial grade. Bleached. Bran particles. Hairs. Total. 15122-K-AA "Per cent clear." 5| 5i 5i 20 25 30 30 30 30 35 50 50 25 243 244 282 137 308 245 208 235 160 247 126 177 253 155 164 99 66 30 167 143 44 40 39 32 68 72 • 398 15122- K-LL. . No 408 15126- K-EEE (?) 381 17178-L-AAS... No 203 17132-L-W No 338 15122- K-DD Yes 412 15122- K-MM No 351 17160- L-D Yes 279 17162-L-U. . . No 200 17133-L-EE.. Yes 286 17167- L-GG... Yes 158 171S6-L-B. No 245 11006-K (?).. 325 26 BULLETIN 839, U. S. DEPARTMENT OF AGRICULTURE. The bran particle count varied from 126 to 308, with an average of 218. The hair count ranged from 30 to 167, with an average of 86. The total offal count ranged from 158 to 412, with an average of 306. CLEAR FLOURS MILLED FROM BLENDED WHEATS. Twelve samples of flour stated to have been milled from blended wheats were examined. The commercial grades ranged from 10 to 30 per cent. Table 30 gives the results of the examination. Table 30. — Results of examination of clear jiours milled from blended wheats . .Sample No. Com- mercial Bleached. Bran particles. Hairs. 115 61 127 65 250 73 297 40 209 96 76 45 55 49 61 47 166 142 112 98 88 44 111 67 Total. 17179-L-ZZ. 17116-L-E.. 17123-L-GG 17125-L-GG 17171-L-C . . 11093-K 11094-K.... 11095-K . . - . 17182-L-II.. 17173-L-Y. . 17172-L-BB 17159-L-W. "Per cent clear. " 10 15 15 15 20 27i 27i 27i 30 40 60 25 No No No No No No Lightly Heavily No No Yes.... No 176 192 323 337 305 121 104 108 308 210 132 178 The bran particle count varied from 55 to 297, with an average count of 139, and the hair count varied from 40 to 142, with an average of 69. The total offal count varied from 104 to 337, with an average of 207. mill streams employed in THE MANUFACTURE OF CERTAIN CLEAR FLOURS. Tables 31 and 32 record the results obtained on certain mill streams which were employed in making up clear flours. As in the case of the commercial grades already considered, these figures are merely submitted to demonstrate the quality of the stocks that might be used in such a flour from the standpoint of offal material. Table 31. — Results of examination of will streams employed in the manufacture of a clear flour (sample No. ni4S-L-BB) milled from hard wheat. Stock. First and third breaks. Second break Third break Fourth break First tailings Second tailings Third tailings Fourth tailings First germ flour Second germ flour First dustings flour.... Third dustings flour... Dust collector material 25 per cent clear flour » Bran particles. Hairs. 310 69 324 42 628 107 810 213 120 5 120 4 38 1 667 72 430 38 560 33 184 28 110 15 575 99 316 71 Total. 360 366 735 1,023 125 124 3» 639 46» 59S 21t 12S 674 38T 1 Composited from the mill streams listed above it. MICROSCOPICAL EXAMINATION OF FLOUR. 27 Table 32. — Results of examination of mill streams employed in the mamifacture of a clear flour (sample No. 11079-K-JJ) milled from hard wheat. Stock. First break Third break (head) , Third break (tail) , Fifth middlings , SLxth middlings (head) Sixth middlings (tail) Seventh middlings (head) . First siziags First tailings (head) , First tailings (tail) 22 per cent clear flour ' Bran particles. Hairs. Total. 196 165 361 120 121 241 100 103 203 28 24 52 46 45 91 55 26 81 56 27 83 87 31 118 151 58 209 87 32 119 82 68 150 1 Composited from the mill streams listed above it. GENERAL CONCLUSIONS ON CLEAR FLOURS. 1. The commercial grades of so-called clear flours ranged from 5^ to 52 per cent. 2. The average total offal count obtained on all commercial clear flours examined was 273. This amount was decidedly in excess of the amount obtained on the commercial grades already considered. 3. As in the case of the other grades, clear flours showed a wide variation in the total offal count obtained on products from different mills. LOW-GRADE FLOURS. The low-grade flour is supposed to be made from low-grade mill stocks, as might be inferred from the designation applied to this class of products. As already stated, the better stocks, for the most part, are diverted into the higher grades. The streams entering into the composition of the low-grade flours are usually more or less specky, due to the presence of offal material. For this reason it is quite impossible to obtain an accurate count on such a flour. In fact, a casual microscopical examination is usually all that is neces- sary to determine the quality of the flour. LOW-GRADE FLOURS MILLED FROM HARD WHEATS. Eleven low-grade flours milled from hard wheats were examined, with the results shown in Table 33. The commercial grades ranged from 2 to 10 per cent, some of the samples being bleached and others unbleached. Table "i'i.^-Results of examination of low-grade flours milled from hard wheats. Sample No. Commer- cial grade. Bleached. Bran particles. Hairs. TotaL 11066-K "Per cent low- grade." 2 2i 2i 2§ 3 2-5 5 5 8 6 10 (?) 243 310 340 310 252 175 353 274 269 169 317 91 129 131 112 155 88 301 335 264 163 238 334 15118-K No 439 15119 K Lightly 471 15120-K Yes 422 a5156-K D No 407 15148- K-X No 263 11080-K HH No 654 1108()-K-II Yes 609 •11029-K D (?) 533 11072 K OO (?) 332 11080- K-OO Yes 555 28 BULLETIN 839, U. S. DEPARTMENT OF AGRICULTURE. The bran particle count varied from 169 to 353, witli an average of 273. The hair count ranged from 88 to 335, with an average of 182. The total offal count varied from 263 to 654, with an average of 456. LOW-GRADE FLOURS MILLED FROM SOFT WHEATS. The eight samples of low-grade flour milled from soft wheats ranged from 2 to 10 per cent, with bleaching being practiced in some instances and not in others. Table 34 gives the results of this examination. Table 34. — Results of examination of loiv-grade flours milled from soft vheatf!. Sample No. Commer- cial grade. Bleached. Bran par- ticles. Hairs. Total. 17136-L-Y "Percent low- grade." ■ (?) (?) 2 3 4i I' 10 (?) 202 113 309 238 402 390 307 331 27 257 1:5 261 219 139 124 80 229 17185-L-G No 409 17176-L-X No 454 17188-L-W No 499 15123-K Z . Yes 621 15126-K-CCC No 529 17178-L-BBS No I.'?! 1716&-L-Y Yes 411 The bran particle count varied from 143 to 402, with an average of 302. The hair count ranged from 27 to 261, with an average of 140. The total offal count varied from 229 to 621, with an average of 446. LOW-GRADE FLOURS MILLED FROM BLENDED WHEATS. Eight samples of flour stated to liave been milled from blended wheats ranged in commercial grades from 1^ to 10 per cent, only one sample of the number being represented as having been bleached. The results of the examination are shown in Table 35. Table 35. — Results of exam,ination of low-grade flours milled from blended wheats. Sample No. Commer- cial grade. Bleached. Bran par- ticles. Hairs. Total. 17123 L EE "Percent low- grade." (?) (?) li 3i 4 5 10 10 No 394 100 211 357 397 237 281 262 59 61 76 141 183 94 131 132 45.1 17128-L-T No 161 17117-L Y No 287 17115- L-F (?) 498 17120- L-J No 580 17171- L-D No 331 17172- L-AA. . Yes 412 17179-I>-AAA No... . . 394 The bran particle count had limitations of from 100 to 397, with an average of 279. The hair count varied from 59 to 183, with an average of 109. The total offal count ranged from 161 to 580. with an avcrajjo count of 389. MICROSCOPICAL EXAMINATION OF FLOUR. 29 GENERAL CONCLUSIONS ON LOW-GRADE FLOURS. 1 . The commercial grades of so-called low-grade flours ranged from 2 to 10 per cent. 2. The average total offal count obtained on all commercial low- grade flours examined was 433. This indicated that not as much attention was given to the purification of the stocks passing into such flours as was done in the case of the stocks composing the grades already considered. 3. The data obtained on the low-grade flours milled from the dif- ferent wheats are summarized in Table 36. Table 36. — Limitations and average counts on bran particles and hairs for low-grade flours. Type. Bran particles. Hairs. Variation. Average. Variation. Average. Hard wheat 169 to 353.... 143 to 402.... 100 to 397.... 273 302 279 88 to 335 27 to 261 59 to 183 182 Soft wheat 140 109 EXAMINATION OF EXPERIMENTAL SERIES OF FLOUR. In connection with the examination of commercial flours it was considered advisable to examine samples of flour whose composition was definitely known, as far as the wheat from which they were milled and their constituent streams were concerned. The information in regard to the commercial samples was definite enough in so far as the milling operator was able to judge. The samples of flour employed in tliis part of the investigation were milled under the personal supervision of B. C. Winslow, food and drug inspector, Bureau of Chemistry, United States Department of Agriculture. The samples were prepared at a plant at Lyons, Kans., a portion being milled from a No. 2 Nebraska hard winter wheat, crop of 1914, containing from 25 to 35 per cent of yellow berry wheat, and another portion from a Kansas No. 2 hard winter wheat. Each type of flour was subjected to three degrees of bleacliing, thus making three samples for each type. Four types of flour were made from each wheat, a 70 per cent, a 90 per cent, a 97.5 per cent, and a 27.5 per cent. In the case of the Kansas wheat a fifth type, a 2.5 per cent, was made. The component streams that passed into each type and the results of the examinations made were as follows: THE 70 PER CENT TYPE OF EXPERIMENTAL FLOUR. COMPOSITION. First sizings floior. Fourth middlings flour. Second sizings flour. Fifth middlings flour. First middlings flour. Fine tailings flour. Second middlings flour. Coarse tailings flour. Third middlings flour. 30 BULLETIN 839, U. S. DEPARTMENT OF AGRICULTURE. Table 37. — Results of examination of 70 per cent type of experimental flour. WTieat. Sample number. Decree of bleaching. Bran particles. Hairs. Total. 111084-K None 29 32 31 10 12 13 13 18 12 9 0) 42 •IllOSS-K Lightly 45 I1IO86-K Heavily 49 (15112-K None 22 •{15113-K Lightly 21 (15114-K Heavily (') Average count 22 13 35 1 Not counted; infested with weevils. THE 90 PER CE>rr TYPE OF EXPERIMENTAL FLOUR. COMPOSITION. First sizings flour. Second sizings flour. First middlings flour. Second middlings flour. Third middlings flour. Fourth middlings flour. Fifth middlings flour. Fine tailings flour. Coarse tailings flour. Second break flour. Third break flovir. Fourth break flour. Sharp section (middlings). Cut-off flour (middlings). Sixth middlings flour. Seventh middlinofs flour. Eighth middlings flour. Table 38. — Results of examination of 90 per cent type experimental flour: Wheat. Sample number. Degree of bleaching. Bran particles . Hairs. ; Total. 1 (11087-K None 50 51 50 32 31 28 26 76 hl088-K Lightly 22 7J Ill089-K Heavily 28 78 nol09-K None 31 0» hsilO-K Lightlv . . 28 34 59 [iSlll-K Heavily 62 40 28 6« THE 97.5 PER CENT TYPE OF EXPERIMENTAL FLOUR. COMPOSITION. First sizings flour. Second sizings flour. First middlings flour. Second middlings flour. Third middlings flour. Fourth break flour. Sharp section (middlings) Cut-oft flour (middlings). Sixth middlings flour. Seventh middlings flour. Fourth middlings flour. Fifth middlings flour. Fine tailings flour. Coarse tailings flour. Second break flour. Third break flour. Eighth middlings flour. First break flour. Fifth break flour. Ninth middlings flour. Flour from dust-collecting reels. Table 3d.— Results of examination of 97.5 per cent type of experimental flour. Wheat. Sample number. Degree of bleaching. Bran particles. Hairs. Total. fll090-K None 43 43 52 57 43 28 28 29 26 39 29 30 70 11091-K Lightlv 72 11092-K Heavily 78 No. 2 Nebraska, hard winter. 1510&-K None 9ft 15107-K Lightly 78 15108-K Heavily 58 44 30 74 MICROSCOPICAL examinatio:n" of flour. 31 THE 27.5 PER CENT TYPE OP EXPERIMENTAL FLOUR. Second break flour. Third break flour. Fourth break flour. Sharp section . COMPOSITION. Cut-off flour (middlings). Sixth middlings flour. Eighth middlings flour. First break flour. Fifth break flour. Ninth break flour. Flour from dust collectors. Seventh middlings flour. Table 40. — Results of examination of 27.5 per cent type of experimental flour. Wheat. Sample number. Degree of bleaching. Bran particles. Hairs. Total. fll093-K None 76 55 61 56 49 51 45 49 47 65 51 40 121 •{11094-K Lightly 104 Ill095-K Heavily 108 (15115-K None..'. 121 No 2 Kansas, hard winter.. h5116-K Lightly 100 [l5117-K Heavily 91 Average 58 49 107 THE 2.5 PER CENT TYPE OF EXPERIMENTAL FLOUR. COMPOSITION. Bran duster flour. Shorts duster flour. Cut-off flour from seventh middlings. Cut-off flour from ninth middlings. Table 41. — Results of examination of 2.5 per cent type of experimental flour . Wheat. Sample number. Degree of bleaching. Bran particles. Hairs. Total. (15118-K None 310 340 310 129 131 112 439 No. 2 Kansas, hard winter... •{15119-K Lightly. . . . 471 Il5120-K Heavily 422 Average 320 124 444 GENERAL CONCLUSIONS ON EXPERIMENTAL TYPES OF FLOUR. The best grade of flour of the experimental series averages a lit- tle lower in total offal count than the best grade in the commercial set, being 57 for the commercial flours and 35 for those of the experi- mental set. The two intermediate grades of the commercial flours were higher in the offal count than similar grades in the experimental series, the count being 111 and 273 for the commercial flours and 71 and 107 for those of the experimental set. Both of the lower- grade flours, that from the commercial and experimental sets, respectively, compared very favorably as far as the offal count was concerned, these figures being essentially minimum ones although approxi- mately representative of the two products. SUMMARY. 1. Microscopical technique was devised for the enumeration of the offal material in flour of various conmiercial grades. 2. The data obtained on the various commercial grades of flour demonstrated that there was little uniformity in the matter of grad- ing finished flours in different mills. 3. The experimental data submitted have shown a wide range in the offal content among flours of the same commercial grade (appar- ently) produced by different mills. 4. The information obtained concerning the samples examined leads to the inference that all mills do not composite finished flours in the same manner. 32 BULLETIN 839, U. S. DEPARTMENT OF AGRICULTURE. 5. The microscopical examination of the constituent streams en- tering into the composition of a finished flour shows the effect of the addition of different mill stocks on the resulting offal content. BIBLIOGRAPHY. (1) Amos, P. A. Processes of Flour Manufacture. 1912. (2) Bellier, J. Recherche microscopique des farines eirangeres dans la farine de bl6, partic- ulierment du riz et de la feverolle. Ann. chim. anal., J^:224-228. 1907. (3) Collin, Eugene. Recherche de la farine de riz dans la farine de h\^. Ann. chim. anal., i./:446- 453. 1906. (4) Dedrick, B. W. Establish Definite Grades of Flour, Offal, and Stock. The Operative Miller, 18:441, 467. 1913. (5) Delate, L. Etude sur I'analyse et I'examen microscopique des farines. Rev. intern, fals., 6:173-175, 188-190. 1893. (6) Gastine, J. Nouveau proc6d6 d'analvse microscopique des farines et recherche du riz dana les farines de hU. Ann. chim. anal., 17:281-283. 1906. (7) GiRARD, Aime. L' Appreciation des farines. La Meunerie Francaise, 12:8-10. 1896. (8) Hanausek, T. F. Ueber unser Mehl und Brot von botanischen Gesichtspunkten betrachtet. Wiener illustrirte Garten-Zeitung, ;g^:109-118. April, 1899. (9) Huss, Harald. Uber die quantitative Bestimmung von vegetabilischen Pulvern mit dem Mikroskop. Landw. Vers. Sta., 60:1-13. 1904. (10) KoHN, Eduard. Beitrage zur Mehluntersuchung. Chem. Ztg., 56:121-123. 1912. (11) Kraemer, H. An Examination of Commercial Flour. J. Am. Chem. Soc, fi:650-663. 1899. (12) Lanoe, Dr. Beitrage zur mikroskop. Untersuchung von Mehlproben. Z. angew. Mik- roskopie, :?:369-370. 1896. (13) Leandre, M. Methodes pratiques d'analvses des farines. Rev. intern, fals., 6:133-136, 153-156. 1893. (14) Marcille, R. Analyse microscopique des farines. Ann. chim. anal., ii:371-372. 1906. (15) Moore, B., and Wilson, J. T. The Effects of Nitrogen Peroxide on the Constituents of Flour in Relation to the Commercial Practice of Bleaching Flour with that Reagent. J. Hyg., 15:438^66. Jan., 1914. (16) POSNER, C. Studien zur Mikroskopie von Mehl und Brot. Zeit. Nahr. Genussm., ^9:329- 337. 1915. (17) Rammstedt, Otto. Die Bestimmung der Farbe des Mehles und das Sichtbarmachen von Klei- teilchen in Mehl und Griesz. Pharm. Zentralhalle, 56:291-293. 1915. (18) Savini, G., and Silvestri, G. Estimation of the Quality of Flour. Ann. Lab. chim. cent. Gabelle., 6:425- 433. 1912. (Abstract in Analyst, 5,^:55-56. 1913.) (19) SCHAFFNIT, E. Zu Mehluntersuchung. Zeit. Nahr. Genussm., 77:86-88. 1909. (20) Snydek, H. The Analysis of WTieat Flour for Commercial Purposes. V. Intemationaler Kongress fiir AngewandteChemie, Section VI, pp. 702-710. Berlin, 1904. (21) Testoni, G. Micro.scopic Analysis of Flour and Bread. Staz. sper. agrar. ital., -^cS": 143-150. 1915. (Abstract in Chem. Abstracts, ,9:2951.) (22) von LiEBERMANN, L., and Andriska, V. Ein neues v'erfahren zur Bestimmung des Feinheitsgrades der Weizeninehle. Zeit. Nahr. Genussm., 22:2%1-2M. 1911. O HBRARY OF CONGRESS 014 338 319 2 \ LIBRARY OF CONGRESS 014 338 319 2 ij^