UNIVERSITY OF 
 
 ILLINOIS LIBRARY 
 
 AT URBANA-CHAMPAIGN 
 
 ACES 
 
NOT/CE: Return or renew all Library Materials. The Minimum Fee for 
 each Lost Book is $50.00. 
 
 The person charging this material is responsible for 
 its return to the library from which >t was withdrawn 
 on or before the Latest Date stamped below. 
 
 '" 
 
 o r 
 
 UNIVERSITY OF ILLINO.S LIBRARY AT URBANA-CHAMPAIGN 
 
 JUN 1 6 2005 
 
 i iMiVERSITY OF ILLINOIS, 
 
UNIVERSITY OF ILLINOIS 
 
 Agricultural Experiment Station. 
 
 URBANA, AUGUST, 1903. 
 
 BULLETIN No. 87. 
 
 THE STRUCTURE OF THE CORN KERNEL 
 
 AND THE COMPOSITION OF ITS 
 
 DIFFERENT PARTS. 
 
 BY CYRIL G. HOPKINS, PH. D., CHIEF IN AGRONOMY AND CHEMISTRY; 
 LOUIE H. SMITH, M. S., CHIEF ASSISTANT IN CHEMISTRY AND PLANT 
 BREEDING; AND EDWARD M. EAST, B. S., ASSISTANT IN CHEMISTRY. 
 
 The possibility of selecting seed corn for improved chemical composi- 
 tion by a simple mechanical examination of sections of kernels (which 
 any one can easily make with a pocket-knife) was clearly established by 
 the experiments reported in Bulletin No. 55, " Improvement in the 
 Chemical Composition of the Corn Kernel" ; and the practical value of 
 this method of selecting seed corn for high protein, high oil, or other de- 
 sirable qualities has been fully confirmed by subsequent investigations 
 and trials by the Experiment Station and by practical seed-corn breeders, 
 as shown in Bulletin No. 82, " Methods of Corn Breeding." 
 
 77 
 
78 
 
 BULLETIN No. 87. 
 
 [August, 
 
 A considerable amount of additional data relating to this matter has 
 been accumulating with the progress of our experiments in corn breed- 
 ing, and because of the very great importance of this subject to the 
 corn growers and corn breeders of Illinois, and also because of the marked 
 
 Stem J 
 
 Embryo) 
 Boot j 
 
 PLATE 1. LOW-PROTEIN CORN KERNEL FROM DRAWING (SMALL KERNELS 
 FROM PHOTOURAPH). 
 
1903.] 
 
 THE STRUCTURE OP THE CORN KERNEL. 
 
 79 
 
 interest which is manifested in this matter both by progressive, prac- 
 tical farmers and by scientific investigators, it has seemed advisable to 
 publish in somewhat greater detail the results of our investigations along 
 this line. 
 
 Horny 
 Gluten 
 
 Horny 
 [Starch 
 
 HmbryoL 
 
 Stem J 
 
 Embryo 
 
 Boob 
 
 Tip 
 Starch 
 
 [Tip 
 
 [Cap 
 
 PLATE 2. HIGH-PROTEIN CORN KERNEL FROM DRAWING (SMALL KERNELS 
 FROM PHOTOGRAPH). 
 
10 BULLETIN No. 87. [August, 
 
 PARTS OF THE CORN KERNEL. 
 
 There are six distinctly different parts in a kernel of corn, as will be 
 readily seen by reference to Plates 1 and 2. 
 
 1. TIP CAP. This is a small cap covering the tip end of the kernel 
 and serves as a protection to the end of the germ. It consists of material 
 somewhat resembling the cob and occasionally in shelling corn the tip 
 cap remains attached to the cob, leaving the tip end of the germ un- 
 covered, but nearly always it remains on the kernel. 
 
 2. HULL. This is the very thin outer covering of the kernel. It con- 
 sists largely of carbohydrates, especially fiber or cellulose, although it 
 also contains a small percentage of other constituents. 
 
 3. HORNY GLUTENOUS PART. This part lies immediately underneath 
 the hull. It constitutes a second covering of the kernel, usually much 
 thicker than the hull. For short, it is called horny gluten, although it 
 is, of course, not pure gluten. However, it is the richest in protein of 
 any part of the corn kernel, as has been stated in bulletins already pub- 
 lished by this station and previously by Doctor Voorhees, Director of the 
 New Jersey Experiment Station. 
 
 4. HORNY STARCHY PART. This part lies next to the horny gluten, 
 on the back and sides of the kernel. For short, it is called homy starch, 
 although it is not pure starch, as it contains considerable amounts of 
 other constituents, especially of protein. In an examination of the 
 kernel with the unaided eye the horny glutenous part and the horny 
 starchy part are not readily distinguished from each other, the line be- 
 tween them being somewhat indefinite and indistinct. Considered both 
 together these two parts constitute the horny part of the kernel. 
 
 5. WHITE STARCHY PART. This part occupies the crown end of the 
 kernel above the germ and it also nearly surrounds the germ toward the 
 tip end of the kernel. For convenience this material is called white 
 starch, although it is not pure starch as has been clearly shown in former 
 publications. In some kernels the horny starch extends nearly or quite 
 to the germ (near the middle of the kernel) and thus separates more or 
 less completely the white starch into two parts which we call crown 
 starch and tip starch. 
 
 6. GERM. The germ occupies the center of the front of the kernel 
 toward the tip end and usually extends about one-half or two-thirds of 
 the length of the kernel. Within the body of the germ are the embryo 
 stem pointing upward toward the crown end and the embryo root point- 
 ing downward toward the tip of the kernel, both of which are, of course, 
 parts of the germ. These embryo parts within the germ may be easily 
 seen by any one who will carefully shave off the front side of the germ 
 from a kernel of corn. (See small photographic reproduction of sections 
 of kernels of high and low protein corn in Plates 1 and 2.) 
 
1903.] THE STRUCTURE OF THE CORN KERNEL. 81 
 
 MECHANICAL SEPARATION OF THE DIFFERENT PARTS. 
 
 It is not a difficult matter to obtain very pure samples of each of the 
 above-named parts of the corn kernel, although in making the separa- 
 tions there is of necessity some waste material consisting of a mixture of 
 three different parts: namely, horny gluten, horny starch, and white 
 starch. 
 
 By the use of a small sharp knife any one can make the following 
 separations : 
 
 1. Tip cap. 
 
 2. Hull. 
 
 3. Horny gluten. 
 
 4. Horny starch.* 
 5a. Crown starch. 
 5b. Tip starch. 
 
 6. Germ. 
 
 7. Waste (mixed material). 
 
 In making these separations the kernels are first soaked in hot water 
 for fifteen or twenty minutes. 
 
 The tip cap is then very easily and perfectly separated by simply 
 cutting under one edge and then pulling it off. 
 
 The hull is separated without difficulty by peeling it off in strips. It 
 is only necessary to use the knife to start the peeling at the tip end where 
 the hull has been broken by removing the tip cap. With some care the 
 hull can be completely peeled out of the dent in the corn. 
 
 The horny gluten is more easily distinguished after the hull is re- 
 moved. It will be plainly seen that it covers the entire kernel, except- 
 ing possibly the exposed part of the germ. The horny gluten is removed 
 by carefully shaving it off with a sharp knife. Adhering particles of 
 starch can be more easily separated from the horny gluten after the 
 shavings have been allowed to dry for some time. In scraping off these 
 particles of horny starch or white starch adhering to the shavings, more 
 or less horny gluten will also be scraped off, so that, while we are thus 
 able to obtain a pure clean sample of the horny gluten, we also obtain 
 some waste material, consisting of particles of horny gluten, horny starch, 
 and white starch. 
 
 The germ is next removed, and with care this can be done very per- 
 fectly. If any particles of starch adhere to the germ they can easily be 
 completely removed. After the tip cap, hull, horny gluten, and germ 
 have been removed, the remainder of the kernel, consisting of the horny 
 starch and white starch only, is allowed to dry, and the kernel is broken 
 in two lengthwise. 
 
 *As used in this bulletin the term "starch" is employed in a technical or com- 
 mercial sense, and not as the name of a definite chemical compound. 
 
82 BULLETIN No. 87. [August, 
 
 The crown starch is dug out with the knife as completely as possible 
 without taking any of the horny starch. 
 
 The tip starch is next removed in the same manner as the crown starch. 
 
 The horny starch from each side usually remains in a solid piece. 
 This is now carefully scraped to remove all adhering particles of white 
 starch or horny gluten, the scrapings being carefully saved and added to 
 the waste material. 
 
 By this method of separation we obtain eight different products, in- 
 cluding the waste material, and seven of these products are pure samples 
 of distinctly different parts of the corn kernel, excepting the crown starch 
 and tip starch, both of which, of course, belong to the white starch; they 
 are kept separate, however, because they are found in different places, 
 frequently being entirely separated in the kernel, although more com- 
 monly there is some white starch continuous from crown to tip. 
 
 COMPOSITION OF THE DIFFERENT PARTS. 
 
 Table 1 shows the percentage of these eight different products, or 
 parts, and the percentage composition of each part, also the percentage 
 composition of the whole corn, for each of three different ears of corn. 
 Ear No. 1 is corn of comparatively low* protein content (see foot-note). 
 Ear No. 2 has about the usual protein content of ordinary corn. Ear 
 No. 3 is high* protein corn (see foot-note). About 200 grams (nearly 
 one-half pound) of kernels from each ear were separated into the differ- 
 ent parts, and each part was then weighed and analyzed separately, 
 another sample of the corn from each ear being analyzed to give the com- 
 position of the whole corn. (All results are given on the water-free basis.) 
 
 *It should be understood that Ear No. 1 (9.28 percent protein) and Ear No. 3 
 (12.85 percent protein) do not represent extremes in protein content; indeed, in 
 our breeding of corn for low protein we have produced good ears containing less 
 than 6.50 percent of protein, and in our high protein field we have produced corn 
 containing over 16 percent of protein. In extremely low protein corn the per- 
 centage of horny part is very much less than in Ear No. 1, and in extremely high 
 protein corn the tip white starch is frequently almost entirely wanting and the crown 
 white starch very greatly reduced, both being replaced by the horny part, as shown 
 in the drawings and also in the actual photographs of sections of kernels shown 
 beside the drawings in Plates 1 and 2. 
 
1903.] 
 
 THE STRUCTURE OF THE CORN KERNEL. 
 
 83 
 
 TABLE 1. PERCENTAGE OF DIFFERENT PARTS AND PERCENTAGE COMPOSITION OF 
 
 EACH PART. 
 
 Ear No. 1 (low in protein). 
 
 Names of parts. 
 
 Percent 
 of 
 whole. 
 
 .Composition of parts. 
 
 Protein, 
 percent. 
 
 Oil, 
 percent. 
 
 Ash, 
 percent. 
 
 Carbo- 
 hydrates, 
 percent. 
 
 Tip caps . 
 
 1.20 
 5.47 
 
 7.75 
 29.58 
 
 16.94 
 10.93 
 
 9.59 
 18.53 
 
 7.36 
 4.97 
 
 19.21 
 8.12 
 
 7.22 
 6.10 
 
 19.91 
 9.90 
 
 9.28 
 
 1.16 
 .92 
 
 4.00 
 .16 
 
 .19 
 .29 
 
 36.54 
 1.06 
 
 4.20 
 
 .91 
 
 .82 
 
 .92 
 .18 
 
 .32 
 .29 
 
 10.48 
 .61 
 
 1.41 
 
 90.57 
 93.29 
 
 75.87 
 91.54 
 
 92.27 
 93.31 
 
 33 .07 
 
 88.43 
 
 85.11 
 
 Hulls 
 
 Horny gluten 
 
 Hornv starch . . . 
 
 Crown starch 
 
 Tip starch 
 
 Germs 
 
 Mixed waste 
 
 Whole corn. . 
 
 Ear No. 2 (medium in protein). 
 
 Tip caps 
 
 1.46 
 
 8.83 
 
 2.30 
 
 1.11 
 
 87.76 
 
 Hulls . 
 
 5 93 
 
 3 96 
 
 89 
 
 79 
 
 94 36 
 
 Horny gluten . 
 
 5 12 
 
 22 50 
 
 6 99 
 
 1 72 
 
 69 09 
 
 Horny starch . 
 
 32 80 
 
 10 20 
 
 24 
 
 24 
 
 89 32 
 
 Crown starch 
 
 11 85 
 
 7 92 
 
 17 
 
 .24 
 
 91.67 
 
 Tip starch 
 
 5 91 
 
 7 68 
 
 39 
 
 .31 
 
 91.62 
 
 Germs 
 
 11.53 
 
 19.80 
 
 34.84 
 
 9.90 
 
 35.46 
 
 Mixed waste.. 
 
 25 40 
 
 11 10 
 
 1 23 
 
 57 
 
 87.10 
 
 Whole corn. . 
 
 
 10.95 
 
 4.33 
 
 1.55 
 
 83.17 
 
 Ear No. 3 (high in protein). 
 
 Tip caps 
 
 1.62 
 
 4.64 
 
 1.99 
 
 1.87 
 
 91.50 
 
 Hulls . .... 
 
 6 09 
 
 3 84 
 
 76 
 
 1.10 
 
 94.30 
 
 Horny gluten 
 
 9.86 
 
 24.58 
 
 4.61 
 
 1.74 
 
 69.07 
 
 Horny starch 
 
 33.79 
 
 10.99 
 
 .22 
 
 .21 
 
 88.58 
 
 Crown starch 
 
 10 45 
 
 8.61 
 
 .52 
 
 .37 
 
 90.50 
 
 Tip starch 
 
 6.23 
 
 7.29 
 
 1.36 
 
 .60 
 
 90.75 
 
 Germs 
 
 11 93 
 
 19.56 
 
 33.71 
 
 10.00 
 
 36.73 
 
 Mixed waste 
 
 20 03 
 
 12.53 
 
 1.15 
 
 .61 
 
 85.71 
 
 Whole corn. . 
 
 
 12.85 
 
 5.36 
 
 1.67 
 
 80.12 
 
 A careful study of Table 1 reveals some very interesting and useful 
 facts regarding the structure of the corn kernel and the composition of 
 the different parts. It is certainly an interesting fact that there are so 
 
84 BULLETIN No. 87. [August, 
 
 many different parts in a kernel of corn, and it is exceedingly useful to be 
 able by a mechanical examination of corn not only to pick out high pro- 
 tein corn or high oil corn as one may desire, but even to separate the 
 several distinctly different parts from one another by purely mechanical 
 means to separate, for example, the horny gluten, containing (in the 
 high protein ear) nearly 25 percent of protein, and then the white 
 starchy parts, with only 7 or 8 percent of protein ; or the germs containing 
 about 35 percent of oil and 10 percent of ash, and then the horny 
 starchy part containing less than 0.25 percent of either oil or ash. 
 
 The hulls contain about 4 percent of protein and are clearly the poorest 
 in protein of any part of the kernel, the next poorest being the tip caps 
 and white starchy parts, containing about 7 or 8 percent, the tip starch 
 being slightly poorer than the crown starch. The horny starch is richer 
 in protein than the white starch, especially in the medium and high pro- 
 tein corn where the difference amounts to more than 2 percent, the 
 horny starchy part containing from 10 to 11 percent of protein. The 
 protein content of the germs is very uniform in the different ears, although 
 the poorest germs are found in the high protein corn, and the richest in 
 the low protein corn, the variation being from 19.56 to 19.91 percent. 
 The horny gluten is the richest in protein of any part of the kernel in both 
 ordinary and high protein corn, as was pointed out several years ago by 
 Doctor Voorhees,* Director of the New Jersey Experiment Station, and 
 as we have quoted in previous publications of the Illinois Experiment 
 Station. In the high protein corn the protein content of the horny 
 gluten amounts to 24.58 percent. In the low protein corn it is slightly 
 less than that of the germ. 
 
 It is plainly seen that the oil in corn is very largely in the germ, al- 
 though the horny gluten also contains a considerable percentage, the germ 
 containing about 35 percent of oil and the horny gluten about 5 per- 
 cent. Both the horny starch and white starch are exceedingly poor in 
 oil, averaging about 0.25 percent, if we disregard the tip starch in Ear 
 No. 3, which appears to have absorbed some oil directly from the germ 
 which it adjoins and partially surrounds. The hulls contain slightly less 
 than 1 percent of oil and the tip caps slightly more than 1 percent, and 
 it is quite possible that this oil may have been obtained, in part at least, 
 by absorption from the horny gluten and germ. Indeed, it seems highly 
 probable that practically all of the true oil in the corn kernel is originally 
 deposited in the germ and horny gluten, and that the small percentage 
 or mere trace, which is found in the other parts is largely obtained by 
 absorption. That such absorption actually does occur is definitely 
 proven by the fact that the percentage of oil in hominy and hominy 
 products increases with the age of the corn used in the milling. (Hominy 
 
 *New Jersey Agricultural Experiment Station Bulletin (1894), 105. 
 
1903.] THE STRUCTURE OF THE CORN KERNEL. 85 
 
 consists largely of the horny starch with more or less adhering white 
 starch.) 
 
 It may be of interest to state in this connection that in 1866 Haber- 
 landt* discovered with the microscope that the germ of the corn kernel 
 contains a large amount of oil. He observed no oil in the remaining por- 
 tions of the kernel. By chemical analysis Lenz* found, however, that 
 after the germs were removed the remaining portion of the kernel con- 
 tained 1.57 per cent of oil. These results were fully confirmed by Doctor 
 Atwaterf who found 1.63 per cent, of oil in the corn after removing the 
 germs and adjoining material, although neither Lenz nor Atwater appear 
 to have ascertained that the horny gluten (the aleurone layer) contains 
 the chief percentage of oil outside of the germ. 
 
 By further reference to Table 1, it will be observed (1) that the germ 
 contains about 10 percent of ash or mineral matter; (2), that this is about 
 ten times the average percentage of ash contained in the other parts ; and 
 (3), that the percentage of ash in the different parts varies with the per- 
 centage of oil, to quite a noticeable degree. 
 
 Of course the percentage of carbohydrates (starch, cellulose, pento- 
 sans, etc.) varies inversely as the sum of the other constituents, being 
 about 35 percent in the germ, 70 percent in the horny gluten, and from 
 90 to 95 percent in the other principal parts. 
 
 The marked degree of uniformity in the entire percentage composition 
 of the germs from each of these three ears, whether low protein, medium 
 protein, or high protein, seems especially noteworthy. The percentage 
 of protein varies only from 19.56 to 19.91; the oil from 33.71 to 36.54; 
 the ash from 9.90 to 10.48; and the carbohydrates from 33.07 to 36.73. 
 It will also be noted that the percentages of both protein and oil are 
 lower in the germs from high protein corn than in those from the low 
 
 protein corn, although the differences are not marked. 
 
 
 
 MATHEMATICAL DISTRIBUTION OF WASTE. 
 
 It will be borne in mind that in making the mechanical separations, 
 in order to obtain each of the seven different parts in pure condition, 
 unmixed with any other part, there was necessarily some waste product. 
 This waste substance amounted to about 20 per cent of the whole. As 
 has already been explained, this mixed waste material consists of only 
 three distinctly different parts horny gluten, horny starch, and white 
 starch (from crown and tip), the other three parts tip caps, hulls, and 
 germs, being easily separated completely and in pure form. 
 
 *Allgempine land und forstwirtschaftliche Zeitung (1866), 257; Jahresbericht 
 (Hoffman) uber die Agricultur-f hemie (1866), 9, 106. 
 
 tThesis, Yale College (1869); American Journal of Science and Arts (1869) (2), 
 48, 352. 
 
86 BULLETIN No. 87. [August, 
 
 By a simple computation the mixed waste material can be distributed 
 among the respective parts of which it is composed, provided we may be 
 allowed to make the assumption (which is approximately the truth) that 
 the horny starch and white starch are present in the waste material in the 
 same proportions as they are in the pure separated portions. Any error 
 which might be introduced by following this assumption would have but 
 little effect because the composition of the horny starch and white starch 
 are not very markedly different (the protein differs by 2 to 3 percent) ; 
 and also because the total amount of waste material to be distributed is 
 only from one-third to one-half the sum of the separated horny starch and 
 white starch. 
 
 It will be observed (see Table 1) that the mixed waste is always richer 
 in protein than the horny starch, thus showing that, besides horny starch 
 and white starch, it also contains more or less horny gluten, which, of 
 course, we know to be the fact. 
 
 If in 100 grams of corn we let 
 x equal the number of grams of tip starch, 
 
 Bx equal the number of grams of crown starch, 
 
 Cx equal the number of grams of horny starch, 
 y equal the number of grams of horny gluten, and 
 <S equal the sum of these four parts, then 
 
 (1) x + Bx + Cx + y=S 
 Now if we let 
 
 a equal the per cent of protein in the tip starch, 
 
 b equal the per cent of protein in the crown starch, 
 
 c equal the per cent of protein in the horny starch, 
 
 d equal the per cent of protein in the horny gluten, and 
 
 s equal the number of grams of protein in all of these four parts, then 
 
 (2) ax + bBx + cCx + dy=s 
 
 Thus we have two equations with which to solve for x and y, which are 
 the only unknown quantities, B and C being factors which can be ob- 
 tained by dividing the percents of separated crown starch and horny 
 starch, respectively, by that of tip starch, and S being the sum of the 
 separated tip starch, crown starch, horny starch, horny gluten and mixed 
 waste, as given in Table 1 ; and a, b, c, d, being the respective percentages 
 of protein in the four separated materials, tip starch, crown starch, horny 
 starch, and horny gluten, and s being the total number of grams of pro- 
 tein contained in these four separated parts and in the mixed waste, all 
 of which data are also given in Table 1. 
 
 PHYSICAL COMPOSITION OF THE CORN KERNEL. 
 
 From the above computations we obtain the results given in Table 2, 
 which gives the total percentages of each of the seven different parts con- 
 
1903.] 
 
 THE STRUCTURE OF THE CORN KERNEL. 
 
 tained in the corn kernel (counting crown starch and tip starch as tw 
 
 parts), and with no waste material. 
 
 TABLE 2. TOTAL PERCENTAGES OF THE DIFFERENT PARTS OF THE CORN KERNEL. 
 
 Name of parts. 
 
 Ear No. 1, 
 (low 
 protein). 
 
 Ear No. 2, 
 (medium 
 protein). 
 
 Ear No. 3, 
 (high 
 protein). 
 
 Tip caps 
 
 1.20 
 
 1 46 
 
 1 62 
 
 Hulls 
 
 5.47 
 
 5 93 
 
 6 09 
 
 Horny gluten 
 
 11 61 
 
 8 51 
 
 13 32 
 
 Horny starch 
 
 37 15 
 
 47 08 
 
 44 89 
 
 Crown starch 
 
 21 26 
 
 17 01 
 
 13 88 
 
 Tip starch 
 
 13 71 
 
 8 48 
 
 6 28 
 
 Germs 
 
 9 59 
 
 11 53 
 
 11 93 
 
 
 
 
 
 Total . 
 
 99 99 
 
 100 00 
 
 100 01 
 
 
 
 
 
 It will be observed that the percentages of horny gluten, horny starch, 
 and germs are noticeably higher in the high protein corn than in the low 
 protein corn; while the opposite is true with the white starch, the per- 
 centages of crown starch and tip starch being markedly higher in the low 
 protein corn than they are in the high protein corn. It is noteworthy 
 that the horny gluten in high protein corn not only contains a higher per- 
 centage of protein than the germs, but that the proportion of horny 
 gluten in the kernel equals or exceeds that of the germs. The only dis- 
 crepancies appearing in Table 2 are the low percentage of horny gluten 
 and the high percentage of horny starch in Ear No. 2. Otherwise the 
 percentages of parts in the medium protein ear are always intermediate 
 between those in the other two ears, as would be expected. Even these 
 discrepancies disappear if the two horny parts be added together and 
 considered as one part, as is done in the practical work of selecting seed 
 corn for higher protein content by mechanical examination, as will be seen 
 by referring to Table 3. 
 
 TABLE 3. PERCENTAGES OF THE DIFFERENT PARTS OF THE CORN KERNEL AS 
 COMMONLY OBSERVED IN MECHANICAL EXAMINATION FOR SEED 
 CORN SELECTION. 
 
 Names of parts. 
 
 Ear No. 1, 
 (low 
 protein). 
 
 Ear No. 2, 
 (medium 
 protein). 
 
 Ear No. 3, 
 (high 
 protein). 
 
 Average 
 percent. 
 
 Tip caps 
 
 1 20 
 
 1 46 
 
 1 62 
 
 1 43 
 
 Hulls 
 
 5 47 
 
 5 93 
 
 6 09 
 
 5 83 
 
 Horny part 
 
 48.76 
 
 55.59 
 
 58.21 
 
 54.19 
 
 White starch 
 
 34.96 
 
 25.49 
 
 22.16 
 
 27.54 
 
 Germs 
 
 9.59 
 
 11.53 
 
 11.93 
 
 11.02 
 
 
 
 
 
 
 Total . . 
 
 99 98 
 
 100 00 
 
 100 01 
 
 100 00 
 
 
 
 
 
 
88 BULLETIN No. 87. [August, 
 
 In this table the crown starch and tip starch are also added together 
 and the sum recorded as white starch. The increase in the amount of 
 horny part (from 48.76 to 58.21 percent), and the decrease in white 
 starch (from 34.96 to 22.16 percent) as we pass from the low protein to 
 the high protein corn, is plainly apparent. 
 
 DISTRIBUTION OF CHEMICAL CONSTITUENTS. 
 
 Table 4 shows the location or complete distribution of the chemical 
 constituents among the seven different physical parts of the corn kernel. 
 In other words, this table represents the separation of 100 grams (or 100 
 pounds) of corn into seven different structural or physical parts, and the 
 subsequent division of each of these parts into the four chemical con- 
 stituents, protein, oil, ash, and carbohydrates. 
 
 The complete data shown in Table 4 are presented especially for the 
 benefit of farmers who are corn breeders, and also for the benefit of 
 the manufacturers of corn products. The agreement between the sum 
 of the separate determinations and the direct determinations of the same 
 constituent in the whole corn is very satisfactory, considering that these 
 results are obtained by computation from the analyses of nine different 
 materials, including the whole corn. The greatest difference is well 
 within the limit of unavoidable error in sampling and analytical deter- 
 minations. A careful study of this table will reveal some interesting and 
 valuable facts. For example, it will be seen that in 100 pounds of the 
 low protein corn the horny gluten contains only 2.23 pounds of protein; 
 while 3.27 pounds of protein are contained in the horny gluten in 100 
 pounds of the high protein corn. Again, in 100 pounds of the low pro- 
 tein corn the horny starch contains only 3.02 pounds of protein; while 
 4.93 pounds of protein are contained in the horny starch in 100 pounds of 
 the high protein corn. 
 
 On the other hand, in 100 pounds of the low protein corn the crown 
 starch and tip starch contain 1.53 and .84 pounds of protein, respectively; 
 while 1.20 and .60 are the respective amounts contained in the correspond- 
 ing parts of the high protein corn. 
 
 If we add together the horny parts and then add together the crown 
 starch and tip starch, as is done in the practical selection of seed corn by 
 mechanical examination, we obtain the results shown in Table 5. 
 
1903.] 
 
 THE STRUCTURE OF THE CORN KERNEL. 
 
 TABLE 4. PHYSICAL AND CHEMICAL DISTRIBUTION OF -100 GRAMS (OR 100 POUNDS) 
 
 OF CORN. 
 
 Ear No. 1 (low in protein). 
 
 Names of parts. 
 
 Physical 
 distribution 
 (grams or 
 pounds). 
 
 Chemical distribution. 
 
 Protein, 
 (grams or 
 pounds). 
 
 Oil, 
 (grams or 
 pounds). 
 
 Ash, 
 (grams or 
 pounds') . 
 
 Carbo- 
 hydrates, 
 (grams or 
 pounds). 
 
 Tip caps . . 
 
 1.20 
 5.47 
 
 11.61 
 37.15 
 
 21.26 
 13.71 
 
 9.59 
 
 .09 
 .27 
 
 2.23 
 3.02 
 
 1.53 
 
 .84 
 
 1.91 
 
 .01 
 .05 
 
 .46 
 .06 
 
 .04 
 .04 
 
 3.50 
 
 .01 
 .04 
 
 .11 
 .07 
 
 .07 
 .04 
 
 1.01 
 
 1.09 
 5.10 
 
 8.81 
 34.01 
 
 19.62 
 12.79 
 
 3.17 
 
 Hulls 
 
 Horny gluten 
 
 Horny starch . . . 
 
 Crown starch 
 
 Tip starch 
 
 Germs 
 
 Total 
 
 99.99 
 
 9.89 
 9.28 
 
 4.16 
 4.20 
 
 1.35 
 1.41 
 
 84.59 
 85.11 
 
 Whole corn. . 
 
 Ear No. 2 (medium in protein). 
 
 Tip caps 
 
 1.46 
 
 .13 
 
 .03 
 
 .02 
 
 1.28 
 
 Hulls 
 
 5.93 
 
 .23 
 
 .05 
 
 .05 
 
 5.60 
 
 Horny gluten 
 
 8 51 
 
 1 89 
 
 .59 
 
 .15 
 
 5.88 
 
 Horny starch 
 
 47.08 
 
 4.80 
 
 .11 
 
 .11 
 
 42.05 
 
 Crown starch . . . . 
 
 17 01 
 
 1 35 
 
 03 
 
 .04 
 
 15.59 
 
 Tip starch 
 
 8.48 
 
 .65 
 
 .03 
 
 .03 
 
 7.77 
 
 Germs 
 
 11 53 
 
 2 28 
 
 4.02 
 
 1.14 
 
 4.09 
 
 
 
 
 
 
 
 Total 
 
 100 00 
 
 11 33 
 
 4 86 
 
 1.54 
 
 82.26 
 
 Whole corn 
 
 
 10 95 
 
 4.33 
 
 1.55 
 
 83.17 
 
 
 
 
 
 
 
 Ear No. 3 (high in protein). 
 
 Tip caps 
 
 1 62 
 
 .08 
 
 .03 
 
 .03 
 
 1.48 
 
 Hulls 
 
 6 09 
 
 23 
 
 .05 
 
 .07 
 
 5.74 
 
 Horny gluten 
 
 13 32 
 
 3 27 
 
 61 
 
 23 
 
 9 20 
 
 Horny starch . 
 
 44 89 
 
 4 93 
 
 10 
 
 09 
 
 39 76 
 
 Crown starch 
 
 13 88 
 
 1 20 
 
 07 
 
 .05 
 
 12.56 
 
 Tip starch ... 
 
 8 28 
 
 60 
 
 11 
 
 .05 
 
 7.51 
 
 Germs 
 
 11 93 
 
 2 33 
 
 4 02 
 
 1 19 
 
 4.38 
 
 
 
 
 
 
 
 Total 
 
 100 01 
 
 12 64 
 
 4 99 
 
 1.71 
 
 80.63 
 
 Whole corn 
 
 
 12.85 
 
 5.36 
 
 1.67 
 
 80.12 
 
 
 
 
 
 
 
90 
 
 BULLETIN No. 87. 
 
 [August, 
 
 TABLE 5. POUNDS OF PROTEIN IN 100 POUNDS OF CORN. 
 
 Names of parts. 
 
 Low 
 protein 
 corn. 
 
 Medium 
 protein 
 corn. 
 
 High 
 protein 
 corn. 
 
 In tip caps . . 
 
 .09 
 
 .13 
 
 .08 
 
 In hulls 
 
 .27 
 
 .23 
 
 .23 
 
 In horny part 
 
 5.25 
 
 6.69 
 
 8.20 
 
 In white starch 
 
 2.37 
 
 2.00 
 
 1.80 
 
 In germs 
 
 1.91 
 
 2.28 
 
 2.33 
 
 
 
 
 
 Total . 
 
 9.89 
 
 11.33 
 
 12.64 
 
 
 
 
 
 It will be observed that the increase in protein in high protein corn 
 over that in low protein corn occurs almost entirely in the horny part of 
 the corn kernel. As indicated in previous bulletins there is also a slight 
 increase in protein in the germ, although this is quite insignificant as com- 
 pared with the increase in the horny part. In passing from low protein 
 corn to high protein corn, there is an appreciable decrease in the amount 
 of protein contained in the white starch. Of course this is due to, the 
 marked decrease in the actual amount of white starch in high protein 
 corn. Indeed, this decrease in the quantity of white starch is even more 
 marked than would appear from Table 5, because the white starch in the 
 high protein corn is actually richer in protein than that in low protein 
 corn, as would be expected and as is shown in Table 1. 
 
 The data given in Table 5 strongly confirm the results which we have 
 already obtained in practical experience in corn breeding. For example, 
 we have been breeding both high protein corn and low protein corn for 
 the past seven years. In the high protein corn we find that the propor- 
 tion of horny part has increased very markedly, while the white starchy 
 part has markedly decreased. In the low protein corn the opposite is 
 true, the horny part having decreased and the white starchy part having 
 markedly increased, in proportion. 
 
 By computation from the data given in Table 4, we have constructed 
 Table 6, which shows the percentage distribution of the different chemical 
 constituents among the several physical parts of the corn kernel. 
 
 It will be seen that as an average about 22 percent of the total protein 
 is contained in the horny gluten, nearly 40 percent in the horny starch, 
 and nearly 20 percent in the germ, thus these three parts contain about 
 80 percent of the total protein in the kernel. 
 
 The germ contains from 80 to 84 percent of the oil, while all other 
 parts combined contain only 15 to 20 percent of the total oil in the 
 kernel. Based upon this fact is the method for selecting high oil or low 
 oil seed corn by mechanical examination, the ears whose kernels show a 
 large proportion of germ being high oil corn and those with small germs 
 
1903.] 
 
 THE STRUCTURE OF THE CORN KERNEL. 
 
 91 
 
 TABLE 6. PERCENTAGE DISTRIBUTION OF CHEMICAL CONSTITUENTS AMONG 
 
 PHYSICAL PARTS. 
 
 Ear No. 1 (low in protein). 
 
 Names of parts. 
 
 Percent 
 of total 
 protein. 
 
 Percent 
 of total 
 oil. 
 
 Percent 
 of total 
 ash. 
 
 Percent 
 of total 
 carbo- 
 hydrates. 
 
 In tip caps . 
 
 .89 
 
 33 
 
 81 
 
 1 29 
 
 In hull ... 
 
 2 75 
 
 1 21 
 
 3 34 
 
 6 03 
 
 In horny gluten.. 
 
 22 56 
 
 11 13 
 
 7 96 
 
 10 41 
 
 In horny starch 
 
 30.51 
 
 1.43 
 
 4 98 
 
 40 22 
 
 In crown starch 
 
 15.52 
 
 .97 
 
 5 07 
 
 23 18 
 
 In tip starch .... 
 
 8.46 
 
 95 
 
 2 96 
 
 15 12 
 
 In germs . . . 
 
 19 31 
 
 83 99 
 
 74 87 
 
 3 75 
 
 
 
 
 
 
 Total 
 
 100.00 
 
 100.01 
 
 99 99 
 
 100 00 
 
 
 
 
 
 
 Ear No. 2 (medium in protein). 
 
 In tip caps 
 
 1.14 
 
 .69 
 
 1.06 
 
 1 56 
 
 In hull 
 
 2.07 
 
 1 08 
 
 3 06 
 
 6 80 
 
 In horny gluten.. 
 
 16 67 
 
 12 21 
 
 9 56 
 
 7 15 
 
 In horny starch 
 
 42.36 
 
 2.32 
 
 7.38 
 
 51.12 
 
 In crown starch.. 
 
 11.88 
 
 .59 
 
 2.67 
 
 18.96 
 
 In tip starch 
 
 5.75 
 
 .68 
 
 1.72 
 
 9.45 
 
 In germs . 
 
 20.14 
 
 82.43 
 
 74.55 
 
 4.97 
 
 
 
 
 
 
 Total .. 
 
 100.01 
 
 100.00 
 
 100.00 
 
 100.01 
 
 Ear No. 3 (high in protein). 
 
 In tip caps . ... 
 
 59 
 
 65 
 
 1 76 
 
 1 84 
 
 In hull 
 
 1.85 
 
 .93 
 
 3.90 
 
 7.12 
 
 In horny gluten 
 
 25 88 
 
 12 29 
 
 13 49 
 
 11.41 
 
 In horny starch 
 
 39 00 
 
 1 98 
 
 5 49 
 
 49.31 
 
 In crown starch 
 
 9 45 
 
 1 44 
 
 2 99 
 
 15 58 
 
 In tip starch 
 
 4 77 
 
 2 25 
 
 2 89 
 
 9 32 
 
 In germs . . . . 
 
 18 45 
 
 80 46 
 
 69 46 
 
 5 43 
 
 
 
 
 
 
 Total . 
 
 99 99 
 
 100 00 
 
 99 98 
 
 100 01 
 
 
 
 
 
 
92 
 
 BULLETIN No. 87. 
 
 [August, 
 
 low oil corn. (See Plate 4.) About 12 percent of the total oil is con- 
 tained in the horny gluten, leaving only about 5 percent of the oil dis- 
 tributed among the remaining five physical parts, and, as stated above, 
 more or less of this small amount is undoubtedly absorbed from the con- 
 tiguous germ or horny gluten. 
 
 It will be noted that the ash is closely associated with the oil, nearly 
 75 percent of the total ash being contained in the germ, and about 10 
 percent in the horny gluten, as an average. 
 
 Table 7 shows, for direct comparison, the percentage distribution of the 
 protein among the different physical parts, in each ear, the two horny 
 parts, and also the two white starchy parts, being combined, as in Table 5. 
 
 TABLE 7. DISTRIBUTION OP 100 GRAMS (OR 100 POUNDS) OF PROTEIN AMONG THE 
 PHYSICAL PARTS AS OBSERVED IN MECHANICAL EXAMINATION. 
 
 Names of parts. 
 
 Low 
 protein 
 Corn. 
 
 Medium 
 protein 
 corn. 
 
 High 
 protein 
 corn. 
 
 In tip caps ... . . 
 
 89 
 
 1 14 
 
 59 
 
 In hulls . 
 
 2 75 
 
 2 07 
 
 1 85 
 
 In horny part 
 
 53 07 
 
 59 03 
 
 64 88 
 
 In white starch 
 
 23 98 
 
 17 63 
 
 14 22 
 
 In germs . . ... 
 
 19 31 
 
 20 14 
 
 18 45 
 
 
 
 
 
 Total 
 
 100 00 
 
 100 01 
 
 99 99 
 
 
 
 
 
 Table 7 illustrates very plainly the fact that, as we pass from low pro- 
 tein corn to high protein corn, the protein decreases in the white starchy 
 part and increases in the horny part; in other words, in breeding corn for 
 high protein, we decrease the white starchy part, which is comparatively 
 poor in protein, and increase the horny part, which averages very much 
 richer in protein, the horny starch containing 2 to 3 percent more pro- 
 tein than the white starch, and the horny gluten being richer in 
 protein than any other part of the kernel. As a rule, in breeding for 
 high protein there is also a slight increase in the proportion of germ, 
 which, being rich in protein, adds somewhat to the increase in protein. 
 
 MECHANICAL METHODS OF SELECTING SEED CORN FOR 
 IMPROVEMENT IN COMPOSITION. 
 
 As has already been shown in our Bulletin No. 82 "Methods of Corn 
 Breeding," we have found it entirely feasible and practical to select seed 
 corn of higher protein content by a simple mechanical examination of a 
 few kernels from each ear. With some care any farmer or corn grower 
 can learn to pick out high protein seed corn by dissecting and examining 
 a few kernels from each ear (by means of a pocket-knife), selecting for 
 high protein seed the ears whose kernels show a large proportion of horny 
 
1903.] THE STRUCTURE OF THE CORN KERNEL. 93 
 
 part, and rejecting those showing a small proportion of horny part. (See 
 Plate 3; also Plates 1 and 2.) 
 
 HIGH PROTEIN KERNELS Low PROTEIN KERNELS 
 
 (Much horny part; (Little horny part; 
 
 little white starch). much white starch). 
 
 PLATE 3. 
 
94 BULLETIN No. 87. [August, 
 
 This method is already in use by practical corn breeders, and with a 
 very satisfactory degree of success. For example, in selecting seed corn 
 by this method Mr. Ralph Allen, of Tazewell County, obtained seed ears 
 for the year 1902, which were 1.46 percent higher in protein than the 
 rejected ears from the same lot, and for this season (1903) his selected 
 seed ears contain 1.58 percent protein more than the ears which he has 
 rejected. In other words, his selected seed corn is richer by 1.58 pounds 
 of protein per 100 pounds of corn than that rejected. 
 
 The method proposed some years ago by Professor Willard, Director of 
 the Kansas Agricultural Experiment Station, of picking out high protein 
 seed by simply selecting for large germs enables one, as a rule, to make 
 some gain in protein, but the gain is very much greater when the propor- 
 tion of horny part is considered. In fact, from our own experience we 
 find that the selection for a large proportion of horny part is a very much 
 more trustworthy index than the size of the germ, in securing high pro- 
 tein seed, and we often find corn with large germs which is actually low 
 in protein, because of a small percentage of protein in the remainder of 
 the kernel. The fact that only 20 percent of the total protein of the 
 kernel is obtained in the germ (as shown in Table 7) is evidence of the 
 uncertainty of obtaining high protein seed corn, and of the improbability 
 of making any very considerable gain in protein, by this method of selec- 
 tion. This difficulty was well understood by Professor Willard, as will 
 be seen in the following quotation from the Kansas Experiment Station 
 Bulletin No. 107, page 63. 
 
 " There are undoubtedly great differences in the protein content of the 
 part of the kernel exclusive of the germ, and it is conceivable and not 
 improbable, that a large germ, though in itself tending to produce high 
 protein content might be overcome by the low protein of the remainder 
 of the kernel."* 
 
 Of course, if one picks out corn with large germs and at the same time, 
 either consciously or unconsciously, selects those ears whose kernels con- 
 tain a large proportion of horny part, he may make considerable gain in 
 protein, but in such case the gain should not be attributed solely to the 
 large germs. 
 
 * Protein is substituted for nitrogen in this quotation. 
 
1903.] 
 
 THE STRUCTURE OF THE CORN KERNEL. 
 
 95 
 
 HIGH OIL KERNELS 
 (Large germs). 
 
 Low OIL KERNELS 
 (Small germs). 
 
 PLATE 4. 
 
 The method of selecting seed corn for high oil content on the basis of 
 large germs (see Plate 4) is certainly well founded, because of the fact that 
 more than 80 percent of the total oil of the kernel is contained in the germ. 
 
96 
 
 BULLETIN No. 87. 
 
 [August, 
 
 THE CORRELATION OF SOME PHYSICAL PARTS AND CHEMICAL 
 CONSTITUENTS OF THE CORN KERNEL. 
 
 As was clearly shown in our Bulletin No. 55, there is usually a marked 
 correlation between the percentage of germ and the percentage of oil in 
 the corn kernel, as will be seen from the following summary of 100 sep- 
 arate determinations reported in that bulletin : 
 
 TABLE No. 8. CORRELATION BETWEEN GERM AND OIL IN CORN KERNELS. 
 
 Number of determinations. 
 
 Small germ corn. 
 
 Large germ corn. 
 
 Germ, 
 percent. 
 
 Oil, 
 percent. 
 
 Germ, 
 percent. 
 
 Oil, 
 percent. 
 
 10 
 
 9 10 
 8 56 
 8.28 
 8.73 
 9 82 
 
 3.58 
 3.22 
 3.64 
 3.32 
 4.30 
 
 14.11 
 12.40 
 12.01 
 13.30 
 11.06 
 
 6.49 
 6.71 
 6.08 
 5.82 
 5.21 
 
 10 
 
 10 
 
 10 
 
 10 
 
 General average ...... 
 
 8.90 
 
 3.61 
 
 12.57 
 
 6.06 
 
 While there is, of course, some variability due in part to the different 
 percentages of oil in germs from different kernels (especially in kernels 
 from different ears) and in part to the different percentages of oil in the 
 horny gluten (and to some extent in other parts) from different ears (see 
 Table 2), nevertheless it will be observed that there is a marked correla- 
 tion between the percentage of germ and the percentage of oil in the corn 
 kernel. In other words, the percentage of oil varies with the percentage 
 of germ. It will be observed, however, that a high percentage of germ is 
 accompanied by a still higher proportionate percentage of oil, indicating 
 that increased proportion of germ in the kernel is due to an increase in the 
 quantity of oil more largely than of the remainder of the germ. 
 
 CORRELATION BETWEEN OIL AND PROTEIN. 
 
 That there is a marked correlation between the percentages of germ 
 and oil in the corn kernel is certainly well established; consequently, if 
 the proportion of germ were a reliable index to the relative protein con- 
 tent of the kernel, there would, of course, be some marked correlation be- 
 tween the percentages of oil and protein in corn. 
 
 In Table 1 of Bulletin No. 55 are recorded the proximate analyses of 
 163 different ears of corn. Table 9 is derived from that data, and shows 
 the percentages of oil and protein contained in the ten ears which are 
 lowest in oil and in the ten ears which are highest in oil, the analyses being 
 arranged in the descending order for protein. 
 
 The correlation between oil and protein is certainly very slight. In 
 the low oil corn the oil varies from only 3.84 to 4.08 percent, while the 
 protein varies from 9.08 to 12.96 percent. In the high oil corn the oil 
 
1903.] 
 
 THE STRUCTURE OF THE CORN KERNEL. 
 
 97 
 
 varies only from 5.46 to 6.02, while the protein varies from 9.58 to 13.87 
 percent, making a difference of 4.29 percent protein between these ex- 
 treme ears, while the difference in oil is only .09 percent between the 
 same ears. The average percentages of oil in these two lots of corn vary 
 from 3.99 to 5.64 percent, while the corresponding averages for protein 
 are 11.02 and 11.19, thus showing very little correlation between the per- 
 centages of oil and protein in the corn kernel. 
 
 TABLE 9. CORRELATION BETWEEN OIL AND PROTEIN IN THE CORN KERNEL. 
 
 Low oil corn. 
 
 High oil corn. 
 
 Oil, 
 
 Protein, 
 
 Oil, 
 
 Protein, 
 
 percent. 
 
 percent. 
 
 percent. 
 
 percent. 
 
 3 97 
 
 12.96 
 
 5.72 
 
 13.87 
 
 3.99 
 
 12.28 
 
 5.51 
 
 13.10 
 
 4 03 
 
 11.71 
 
 5 61 
 
 12.09 
 
 4.07 
 
 11.49 
 
 5.75 
 
 11.18 
 
 3.84 
 
 11.29 
 
 5 65 
 
 10.82 
 
 4 08 
 
 11.29 
 
 5.51 
 
 10 49 
 
 3.94 
 
 10.97 
 
 5.46 
 
 10.32 
 
 4.01 
 
 9.68 
 
 5.51 
 
 10.23 
 
 3 95 
 
 9.44 
 
 6 02 
 
 10.18 
 
 4.05 
 
 9.08 
 
 5 63 
 
 9.58 
 
 AveragesS . 99 
 
 11.02 
 
 5.64 
 
 11.19 
 
 Table 10 is also derived from Table 1 of Bulletin No. 55, and is similar 
 to Table 9, except that the analyses of the ten ears lowest in protein and 
 of the ten ears highest in protein are chosen. This shows the reverse 
 correlation ; that is, between protein and oil. Of course, the results should 
 be practically the same as shown in Table 9, the chief value of Table 10 
 being that it uses almost an entirely different set of analyses, and con- 
 sequently gives a duplicate illustration of the lack of correlation between 
 these two constituents. 
 
 TABLE 10. CORRELATION BETWEEN PROTEIN AND OIL IN THE CORN KERNEL. 
 
 Low protein corn. 
 
 High protein corn. 
 
 Protein, 
 
 Oil, 
 
 Protein, 
 
 Oil, 
 
 percent. 
 
 percent. 
 
 percent. 
 
 percent. 
 
 9 31 
 
 4.96 
 
 13 87 
 
 5.72 
 
 8 40 
 
 4.91 
 
 13 10 
 
 5.51 
 
 8 38 
 
 4.88 
 
 12 68 
 
 5.29 
 
 9 31 
 
 4.82 
 
 12 81 
 
 5.21 
 
 8 25 
 
 4.81 
 
 12 63 
 
 5.15 
 
 9 22 
 
 4.60 
 
 13 06 
 
 4.93 
 
 9 15 
 
 4.55 
 
 12 57 
 
 4.82 
 
 9 30 
 
 4.38 
 
 12 79 
 
 4.25 
 
 9 12 
 
 4.10 
 
 12 76 
 
 4.10 
 
 9 08 
 
 4.05 
 
 12.96 
 
 3.97 
 
 Averages... 8. 95 
 
 4.61 
 
 12.92 
 
 4.90 
 
BULLETIN No. 87. 
 
 [August, 
 
 It will be observed that, although the averages for protein vary from 
 8.95 to 12.92 percent, the averages for oil vary only from 4.61 to 4.90 per- 
 cent. These averages indicate only a slight correlation between protein 
 and oil, and the analyses of the individual ears show that even this cor- 
 relation is by no means constant. 
 
 In connection with some investigations relative to heredity Professor 
 Frank Smith of this University, has prepared correlation tables (see 
 Tables 11 and 12) from the 163 analyses of individual ears of corn recorded 
 in Table 1 of Bulletin No. 55. Table 11 shows the correlation between 
 
 TABLE 11. ONE HUNDRED SIXTY-THREE EARS OF CORN GROUPED ACCORDING TO 
 PERCENTAGES OF PROTEIN AND CARBOHYDRATES. 
 
 Carbo- 
 hydrates 
 percent. 
 
 Protein, percent. 
 
 Total 
 No. of 
 ears. 
 
 8 
 
 8.5 
 
 9 
 
 9.5 
 
 10 
 
 10 5 
 
 II 
 
 11.5 
 
 12 
 
 12.5 
 
 13 
 
 13 5 
 
 14 
 
 79 
 
 
 
 
 
 
 
 
 
 
 
 
 
 1 
 
 1 
 
 79 5 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 80 
 
 
 
 
 
 
 
 
 
 
 
 1 
 
 
 
 1 
 
 80.5 
 
 
 
 
 
 
 
 
 
 
 1 
 
 2 
 
 
 
 3 
 
 81 
 
 
 
 
 
 
 
 
 
 1 
 
 3 
 
 
 
 
 4 
 
 81.5 
 
 
 
 
 
 
 
 1 
 
 
 2 
 
 7 
 
 3 
 
 
 
 13 
 
 82 
 
 
 
 
 
 
 
 2 
 
 5 
 
 13 
 
 
 
 
 
 20 
 
 82.5 
 
 
 
 
 
 1 
 
 3 
 
 7 
 
 12 
 
 3 
 
 1 
 
 
 
 
 27 
 
 83 
 
 
 
 
 
 2 
 
 7 
 
 16 
 
 7 
 
 
 
 
 
 
 32 
 
 83.5 
 
 
 
 
 2 
 
 7 
 
 9 
 
 4 
 
 1 
 
 
 
 
 
 
 23 
 
 84 
 
 
 
 
 3 
 
 7 
 
 5 
 
 1 
 
 
 
 
 
 
 
 16 
 
 84 5 
 
 
 
 1 
 
 8 
 
 2 
 
 
 
 
 
 
 
 
 
 11 
 
 85 
 
 
 
 1 
 
 5 
 
 
 
 
 
 
 
 
 
 
 6 
 
 85.5 
 
 1 
 
 2 
 
 2 
 
 1 
 
 19 
 
 
 
 
 
 
 
 
 
 6 
 
 Total 
 No. of 
 ears. 
 
 1 
 
 2 
 
 4 
 
 19 
 
 24 
 
 31 
 
 25 
 
 19 
 
 12 
 
 6 
 
 
 1 
 
 163 
 
 The correlation of high protein with low carbohydrates and vice versa equals 
 90.05 per cent. 
 
 protein and carbohydrates, the entire 163 ears of corn being grouped ac- 
 cording to the percentage of protein and carbodydrates which they con- 
 tain. This table well illustrates what is meant by a high degree of cor- 
 relation. With an increasing protein percentage there is a decreasing 
 percentage of carbohydrates, and there are no marked exceptions to this 
 rule. Note, for example, that there is one ear containing 14 percent of 
 
1903.] 
 
 THE STRUCTURE OF THE CORN KERNEL. 
 
 99 
 
 protein, but that this ear contains only 79 percent of carbohydrates, 
 being the highest in protein and also the lowest in carbohydrates of the 
 163 ears. There are nineteen ears containing only 12 percent of protein, 
 but none of these nineteen ears contains less than 81 percent of carbo- 
 hydrates; thirteen of them contain 82 percent of carbohydrates, while 
 five others are within one-half percent of that amount. One ear con- 
 tains only 8 percent of protein, but this ear contains 85.5 percent of 
 carbohydrates, being both the lowest in protein and the highest in carbo- 
 hydrates of the 163 ears. It will be observed that of the 163 ears, several 
 groups fall in the squares representing low protein (8 to 10 percent) and 
 high carbohydrates (84 to 85.5 percent) ; also that many ears fall in the 
 squares representing high protein (12 to 14 percent) and low carbo- 
 hydrates (80 to 82 percent) ; so that, including the ears with medium con- 
 tent of protein and carbohydrates, we find that the group numbers of 163 
 ears fall almost in a straight line extending from the lower left-hand cor- 
 ner to the upper right-hand corner of the table. It will be seen that no 
 ears whatever fall in the squares representing high protein and high car- 
 bohydrates, or in those representing low protein and low carbohydrates. 
 By mathematical computation Professor Smith has found that we have 
 in this table 90.05 percent of a perfect correlation. 
 
 TABLE 12. ONE HUNDRED SIXTY-THREE EARS OF CORN GROUPED ACCORDING TO 
 PERCENTAGES OF PROTEIN AND OIL. 
 
 Oil, 
 percent. 
 
 Protein, percent. 
 
 Total 
 No. of 
 ears. 
 
 8 
 
 85 
 
 9 
 
 95 
 
 10 
 
 10 5 
 
 II 
 
 115 
 
 12 
 
 12 5 
 
 13 
 
 13.5 
 
 14 
 
 38 
 
 
 
 
 
 
 
 
 1 
 
 
 
 
 
 
 1 
 
 40 
 
 
 
 2 
 
 2 
 
 
 
 2 
 
 3 
 
 1 
 
 1 
 
 1 
 
 
 
 12 
 
 4.2 
 
 
 
 
 3 
 
 2 
 
 3 
 
 3 
 
 2 
 
 3 
 
 
 2 
 
 
 
 18 
 
 44 
 
 
 
 
 8 
 
 4 
 
 2 
 
 7 
 
 3 
 
 2 
 
 
 
 
 
 21 
 
 46 
 
 
 
 2 
 
 1 
 
 1 
 
 8 
 
 5 
 
 8 
 
 4 
 
 3 
 
 
 
 
 32 
 
 48 
 
 1 
 
 1 
 
 
 5 
 
 6 
 
 3 
 
 7 
 
 2 
 
 4 
 
 5 
 
 
 
 
 34 
 
 50 
 
 
 1 
 
 
 2 
 
 1 
 
 3 
 
 3 
 
 5 
 
 4 
 
 
 1 
 
 
 
 20 
 
 52 
 
 
 
 
 2 
 
 3 
 
 2 
 
 2 
 
 1 
 
 
 3 
 
 1 
 
 
 
 14 
 
 5.4 
 
 
 
 
 
 
 2 
 
 
 
 
 
 
 
 
 2 
 
 5.6 
 
 
 
 
 1 
 
 1 
 
 1 
 
 1 
 
 
 1 
 
 
 1 
 
 
 
 6 
 
 58 
 
 
 
 
 
 
 
 1 
 
 
 
 
 
 
 1 
 
 2 
 
 6.0 
 
 
 
 
 
 1 
 
 
 
 
 
 
 
 
 
 1 
 
 Total 
 No. of 
 ears. 
 
 1 
 
 2 
 
 4 
 
 19 
 
 19 
 
 24 
 
 31 
 
 25 
 
 19 
 
 12 
 
 6 
 
 
 1 
 
 163 
 
 The correlation of high protein with high oil and of low protein with low oil 
 equals 3.81 percent. 
 
100 BULLETIN No. 87. [August, 
 
 Table 12 shows the degree of correlation which exists between oil and 
 protein. It will be seen that there is no such grouping as in Table 11. 
 In other words, there is no marked correlation between the oil and pro- 
 tein in corn. Some ears are rich in one of these constituents and poor 
 in the other; other ears are rich in both, and still others are poor in both, 
 so that the grouping of the numbers of ears according to the percentages 
 of oil and protein which they contain resembles a circle much more nearly 
 than a straight line. By computation it is found that there is a slight 
 tendency for the protein to increase with increasing oil content, but the 
 degree of correlation amounts to only 3.81 percent of a perfect correlation. 
 
 All of the above data tend to prove that, as the percentage of protein 
 increases in corn the starch decreases, while the oil remains almost un- 
 changed; and that we may increase or decrease the percentage of oil or of 
 germ in corn without markedly affecting the percentage of protein. This 
 was the conclusion drawn when the above-mentioned 163 ears of corn 
 were analyzed more than six years ago. The different strains of corn 
 which we have finally produced in our regular corn breeding work fur- 
 nish us excellent material for ascertaining what effect is produced upon 
 the oil content of corn by breeding for a higher or lower protein content; 
 and, vice versa, what effect is produced upon the protein content by breed- 
 ing for a higher or lower oil content. 
 
 In 1900 we planted ten field rows (called the "mixed plot") with two 
 kinds of corn in every row, one kind having been bred for four years for 
 high oil content, the other (originally from the same variety and stock) 
 having been bred during the same four years for low oil content. These 
 two kinds of seed were planted in every row and in fact in every hill, 
 the low oil kernels and the high oil kernels in the same hill just far 
 enough apart so that the identity of the individual plants could be known 
 as they grew during the season. The corn from each of the ten rows was 
 harvested in two lots, one lot being corn from high oil seed and the other 
 lot being from low oil seed. The two lots from each row were kept sep- 
 arate, the one being labeled as corn from the " high oil side" of the row 
 and the other from the " low oil side." 
 
 The percentages of oil and of protein contained in these different lots 
 of corn are shown in Table 13. 
 
 These data are considered very reliable, both kinds of corn having been 
 grown during the same season and in exactly the same soil and each in- 
 dividual sample whose composition is shown in Table 13 being a com- 
 posite sample representing many ears. The average difference in oil 
 content between the high oil side and the low oil side is 1.97 percent of 
 oil, while the average difference in protein is 0.18 percent. Considering 
 that the percentage of protein in the corn is twice as large as the per- 
 centage of oil, it will be seen that there is less than 5 percent of a perfect 
 correlation between the oil and protein. 
 
1903.] 
 
 THE STRUCTURE OF THE CORN KERNEL. 
 
 101 
 
 TABLE 13. OIL AND PROTEIN IN CORN HARVESTED FROM THE " MIXED OIL" 
 
 PLOT IN 1900. 
 
 
 Low oil side. 
 
 High oil side. 
 
 Row No. 
 
 Oil, 
 
 Protein, 
 
 Oil, 
 
 Protein, 
 
 
 percent. 
 
 percent. 
 
 percent. 
 
 percent. 
 
 1 
 
 3.93 
 
 10.07 
 
 5.61 
 
 10.06 
 
 2 
 
 3 78 
 
 9.26 
 
 5.74 
 
 9.05 
 
 3 
 
 3 73 
 
 10.21 
 
 5 88 
 
 9.12 
 
 4 
 
 3 75 
 
 8.47 
 
 5 99 
 
 9.65 
 
 5 
 
 3.89 
 
 9.39 
 
 5 71 
 
 10.08 
 
 6 
 
 3 80 
 
 9.77 
 
 5 91 
 
 10.23 
 
 7 
 
 3 60 
 
 9.80 
 
 5 60 
 
 9.91 
 
 8 
 
 3 58 
 
 9.65 
 
 5.84 
 
 10.32 
 
 9 
 
 4 22 
 
 9.18 
 
 5.68 
 
 9.15 
 
 10 
 
 3.27 
 
 9.26 
 
 5.82 
 
 9.32 
 
 Average 
 
 3 81 
 
 9.51 
 
 5.78 
 
 9.69 
 
 COMPOSITION OF PEDIGREED CORN. 
 
 In order that it might be shown with even more absolute certainty 
 which physical part of the corn kernel should be increased in order to in- 
 crease the protein content, or the oil content, etc., forty ears of corn were 
 selected from the 1902 crop from our oldest breeding plots, ten ears being 
 taken from each of four different strains, namely : 
 
 " Illinois" High Protein Corn. 
 
 " Illinois" Low Protein Corn. 
 
 "Illinois" High Oil Corn. 
 
 "Illinois" Low Oil Corn. 
 
 each of which represents the seventh generation of pedigreed corn, bred 
 as indicated by the name. Twenty-five average kernels were taken from 
 each of these forty ears, the germs separated from the remainder of the 
 kernel and both parts (that is, germs and endosperms*) analyzed sepa- 
 rately, for each ear, another sample of the corn from each ear also being 
 analyzed to show the composition of the whole corn. The tabular state- 
 ments show the results obtained. 
 
 By referring to Tables 14 and 15 it will be seen that the protein con- 
 tent of the low protein ears varies from 6.36 to 7.09 percent, with an aver- 
 age of 6.71 percent, while the protein content of the high protein ears 
 varies from 13.98 to 15.01, with an average of 14.44 percent. The 
 average oil content of the low protein corn is 4.21 percent and of the 
 high protein ears 4.93 percent. The general averages indicate a slight 
 correlation between oil and protein; however, there are several of the 
 high protein ears which contain less oil than some of the low protein ears, 
 thus showing that such correlation is not constant. 
 
 *As here used the term endosperm includes all parts of the kernel except the 
 germ. 
 
102 
 
 BULLETIN No. 87. 
 
 [August, 
 
 TABLE 14. CHEMICAL COMPOSITION OF TEN EARS OF 
 PROTEIN CORN. 
 
 ILLINOIS " Low 
 
 Ear 
 No. 
 
 Protein, 
 percent. 
 
 Oil, 
 percent. 
 
 Ash, 
 percent. 
 
 Carbohydrates, 
 percent. 
 
 4276 
 
 6.98 
 
 4.69 
 
 1.43 
 
 86.90 
 
 4281 
 
 6.60 
 
 4.21 
 
 1.33 
 
 87.86 
 
 4286 
 
 6.87 
 
 4.16 
 
 1.19 
 
 87.78 
 
 4287 
 
 6.37 
 
 3.90 
 
 1.40 
 
 88.33 
 
 4295 
 
 6.46 
 
 4.14 
 
 1.15 
 
 88.25 
 
 4313 
 
 7.01 
 
 4.13 
 
 1.27 
 
 87.59 
 
 4321 
 
 7.09 
 
 3.84 
 
 1.46 
 
 87.61 
 
 4328 
 
 6.36 
 
 4.05 
 
 1.43 
 
 88.16 
 
 4346 
 
 6.89 
 
 4.33 
 
 1.45 
 
 87.33 
 
 4368 
 
 6.48 
 
 4.67 
 
 1.56 
 
 87.29 
 
 Average . 
 
 6.71 
 
 4.21 
 
 1.37 
 
 87.71 
 
 TABLE 15. CHEMICAL COMPOSITION OF TEN EARS OF "ILLINOIS' 
 PROTEIN CORN. 
 
 HIGH 
 
 Ear 
 
 Protein, 
 
 Oil, 
 
 Ash, 
 
 Carbohydrates, 
 
 No. 
 
 percent. 
 
 percent. 
 
 percent. 
 
 percent. 
 
 4174 
 
 14.70 
 
 5.87 
 
 1.48 
 
 77.95 
 
 4189 
 
 14.74 
 
 4.46 
 
 1.70 
 
 79.10 
 
 4202 
 
 14.21 
 
 4.54 
 
 1.50 
 
 79.75 
 
 4212 
 
 14.61 
 
 4.57 
 
 1.66 
 
 79.16 
 
 4218 
 
 14.37 
 
 4.61 
 
 1.63 
 
 79.39 
 
 4227 
 
 14.03 
 
 5.26 
 
 1.54 
 
 79.17 
 
 4242 
 
 14.28 
 
 5.33 
 
 1.51 
 
 78.88 
 
 4244 
 
 14.49 
 
 4.71 
 
 1.57 
 
 79.23 
 
 4253 
 
 15.01 
 
 5.01 
 
 1.38 
 
 78.60 
 
 4265 
 
 13.98 
 
 4.96 
 
 1.67 
 
 79.39 
 
 Average . 
 
 14.44 
 
 4.93 
 
 1.56 
 
 79.06 
 
 Table 16 shows the percentages of protein and of germ in both the low 
 protein and high protein ears. 
 
 TABLE 16. PROTEIN AND GERM IN Low PROTEIN AND HIGH PROTEIN CORN. 
 
 Low protein corn. 
 
 High protein corn. 
 
 Ear 
 No. 
 
 Protein, 
 percent. 
 
 Germ, 
 percent. 
 
 Ear 
 No. 
 
 Protein, 
 percent. 
 
 Germ, 
 percent. 
 
 4276 
 
 6 98 
 
 9.52 
 
 4174 
 
 14.70 
 
 13.31 
 
 4281 
 
 6 60 
 
 8.81 
 
 4189 
 
 14 74 
 
 9.51 
 
 4286 
 
 6 87 
 
 8.42 
 
 4202 
 
 14 21 
 
 11.44 
 
 4287 
 
 6 37 
 
 9.53 
 
 4212 
 
 14.61 
 
 10.92 
 
 4295 
 
 6.46 
 
 7.95 
 
 4218 
 
 14.37 
 
 11.64 
 
 4313 
 
 7 01 
 
 8.98 
 
 4227 
 
 14 03 
 
 11.15 
 
 4321 
 
 7.09 
 
 10.30 
 
 4242 
 
 14.28 
 
 13.21 
 
 4328 
 
 6.36 
 
 8.88 
 
 4244 
 
 14.49 
 
 11.22 
 
 4346 
 
 6 89 
 
 10.14 
 
 4253 
 
 15 01 
 
 9.82 
 
 4368 
 
 6 48 
 
 10.79 
 
 4265 
 
 13.98 
 
 12.14 
 
 Average . . . 
 
 6 71 
 
 9.33 
 
 Average . . 
 
 14.44 
 
 11.44 
 
1903.] 
 
 THE STRUCTURE OF THE CORN KERNEL. 
 
 103 
 
 Here again we see a slight correlation between the average percentages 
 of protein and germ in the corn kernel, although there are noteworthy 
 discrepancies. Thus we have an ear containing 6.37 percent of protein 
 and 9.53 percent of germ, while another ear contains 14.74 percent of 
 protein and 9.51 percent of germ. In other words, the two ears contain 
 practically the same percentage of germ, although one of them contains 
 more than twice as much protein as the other. One of the lowest pro- 
 tein ears (6.48) contains 10.79 percent of germ, while the highest pro- 
 tein ear (15.01 percent) contains only 9.82 percent of germ. 
 
 Attention is called to the fact that in selecting seed corn by chemical 
 analysis for high protein there is a tendency to increase not only the horny 
 starchy part (which contains more total protein than any other part of the 
 corn kernel), but also to increase both the horny gluten and the germ, 
 both of which, although small in amount are rich in protein; and con- 
 sequently there is a slight tendency for the oil to be increased, not only in 
 the germ, but also in the horny gluten (aleurone layer), which it will be 
 remembered is also quite rich in oil. This is the evident explanation as 
 to why there is a slightly higher degree of correlation between oil and 
 protein in our pedigreed strains of corn than there is in ordinary corn 
 which has not been so bred. 
 
 Tables 17 and 18 show the percentage composition of the low oil and 
 high oil ears. 
 
 TABLE 17. CHEMICAL COMPOSITION OF TEN EARS OF " ILLINOIS" Low OIL CORN. 
 
 Ear 
 No. 
 
 Protein, 
 percent. 
 
 Oil, 
 percent. 
 
 Ash, 
 percent. 
 
 Carbohydrates, 
 percent. 
 
 4474 
 
 9.40 
 
 2 68 
 
 1.45 
 
 86.47 
 
 4486 
 
 9.16 
 
 2 65 
 
 1.64 
 
 86.55 
 
 4491 
 
 9.49 
 
 2 60 
 
 1.29 
 
 86.62 
 
 4495 
 
 9.57 
 
 2.59 
 
 1.41 
 
 86.43 
 
 4509 
 
 8.96 
 
 2.53 
 
 1.36 
 
 87.15 
 
 4512 
 
 10.64 
 
 2 45 
 
 1.46 
 
 85.45 
 
 4521 
 
 9.97 
 
 2 12 
 
 1.42 
 
 86.49 
 
 4537 
 
 10.89 
 
 2 40 
 
 1.54 
 
 85.17 
 
 4548 
 
 9.77 
 
 2.54 
 
 1.36 
 
 86.33 
 
 4555 
 
 11.92 
 
 2.65 
 
 1.42 
 
 84.01 
 
 Average . 
 
 9.98 
 
 2 52 
 
 1.44 
 
 86.07 
 
 The average for the low oil corn is 2.52 percent of oil and 9.98 percent 
 of protein, while the high oil contains 7.00 percent of oil and 11.31 percent 
 of protein. In other words, the high oil corn contains almost three times 
 as much oil as the low oil corn, but is less than one-seventh richer in pro- 
 tein, showing only slight correlation between oil and protein, and with 
 several ears no correlation whatever exists. For example, we have one 
 ear with 10.89 percent of protein and 2.40 percent of oil and another 
 ear with 10.79 percent of protein and 7.01 percent of oil, the protein 
 
104 
 
 BULLETIN No. 87. 
 
 [August, 
 
 being practically equal, while the one ear contains nearly three times as 
 much oil as the other. Again the low oil ear No. 4555 (2.65 percent of 
 oil) contains 11.92 percent of protein, or .61 percent more than the 
 average of all of the high oil ears. 
 
 TABLE 18. CHEMICAL COMPOSITION OF TEN EARS OP " ILLINOIS" HIGH OIL CORN. 
 
 Ear 
 
 No. 
 
 Protein, 
 percent. 
 
 Oil, 
 percent. 
 
 Ash, 
 percent. 
 
 Carbohydrates, 
 percent. 
 
 4374 
 
 11.26 
 
 7 10 
 
 1.64 
 
 80.00 
 
 4411 
 
 10.79 
 
 7.01 
 
 1.40 
 
 80.80 
 
 4412 
 
 9.58 
 
 6 87 
 
 1.58 
 
 81.97 
 
 4417 
 
 10.33 
 
 7 01 
 
 1.66 
 
 81.00 
 
 4421 
 
 12.55 
 
 7.02 
 
 1.53 
 
 78.90 
 
 4423 
 
 11.66 
 
 6 95 
 
 1.56 
 
 79.83 
 
 4436 
 
 11.47 
 
 7 17 
 
 1.59 
 
 79.77 
 
 4441 
 
 12.94 
 
 7 37 
 
 1.57 
 
 78.12 
 
 4448 
 
 11.75 
 
 6 78 
 
 1.52 
 
 79.95 
 
 4462 
 
 10.76 
 
 6.74 
 
 1.48 
 
 81.02 
 
 Average . 
 
 11.31 
 
 7.00 
 
 1.55 
 
 80.14 
 
 TABLE 19. OIL AND GERM IN Low OIL AND HIGH OIL CORN. 
 
 Low oil corn. 
 
 High oil corn. 
 
 Ear 
 No. 
 
 Oil, 
 
 percent. 
 
 ' Germ, 
 percent. 
 
 Ear 
 No. 
 
 Oil, 
 
 percent. 
 
 Germ, 
 percent. 
 
 4474 
 4486 
 4491 
 4495 
 4509 
 
 2.68 
 2.65 
 2 60 
 2.59 
 2.53 
 
 8.05 
 8.13 
 7.92 
 7.39 
 7.06 
 
 4374 
 4411 
 4412 
 4417 
 4421 
 
 7.10 
 7 01 
 6 87 
 7 01 
 7.02 
 
 12.90 
 12.73 
 13.73 
 14.50 
 14.65 
 
 4512 
 4521 
 4537 
 4548 
 4555 
 
 Average . . 
 
 2.45 
 2.12 
 2 40 
 2.54 
 2.65 
 
 7.89 
 7.13 
 7.57 
 7.83 
 8.47 
 
 4423 
 4436 
 4441 
 4448 
 4462 
 
 Average . . 
 
 6.95 
 7.17 
 7 37 
 6 78 
 6.74 
 
 13.83 
 14.10 
 14.53 
 
 14.35 
 13.03 
 
 2 52 
 
 7.74 
 
 7 00 
 
 13.84 
 
 Table 19, giving the percentages of oil and germ in the low oil and 
 high oil corn, shows a very marked correlation between oil and germ. 
 The ten low oil ears contain from 2.12 to 2.68 percent of oil (average 
 2.52) and from 7.06 to 8.47 percent of germ (average 7.74), while the 
 ten high oil ears contain from 6.74 to 7.37 percent of oil (average 7.00), 
 and from 12.73 to 14.65 percent of germ (average 13.84). There is no 
 overlapping and the correlation is very distinct. Every low oil ear con- 
 tains a small percentage of germ and every high oil ear a high percentage 
 of germ. Attention is called to the fact that the high oil corn is even 
 richer in oil than would be indicated by the high percentage of germ as 
 compared with the percentage of oil and germ in the low oil corn, indicat- 
 ing that the breeding for high oil has not only increased the oil by in- 
 
1903.] 
 
 THE STRUCTURE OF THE CORN KERNEL. 
 
 105 
 
 creasing the percentage of germ (which contains most of the oil), but that 
 the percentage of oil in the germ itself has increased. (Of course there 
 is also an increase in the percentage of oil in the horny glutenous part.) 
 Similarly the percentage of oil has decreased even more rapidly than the 
 percentage of germ in the low oil corn. These results are very apparent 
 in the data shown in Table 20. 
 
 EFFECT OF BREEDING ON COMPOSITION OF GERMS AND ENDOSPERMS. 
 As already explained, ten ears were selected from each of the four 
 different strains of corn (low protein, high protein, low oil, and high oil), 
 and twenty-five kernels were taken from each of these forty ears, the 
 germ being separated from the remainder of the kernel, which we call 
 the endosperm. After the percentage of germ was determined from each 
 individual ear, the germs from each lot of ten ears were put together to 
 make two samples, each sample representing five ears. The endosperms 
 were likewise put together, so that we have duplicate samples of both 
 germs and endosperms for each of the four different strains. These 
 samples were analyzed chemically and the results are given in Table 20. 
 
 TABLE 20. CHEMICAL COMPOSITION OF GERMS AND ENDOSPERMS FROM Low PRO- 
 TEIN AND HIGH PROTEIN CORN AND FROM Low OIL 
 AND HIGH OIL CORN. 
 
 Kind of 
 corn. 
 
 Part of 
 kernel. 
 
 Protein, 
 percent. 
 
 Oil, 
 
 percent. 
 
 Ash, 
 percent. 
 
 Carbo- 
 hydrates, 
 percent. 
 
 Low protein 
 
 Germs 
 
 ( 18.05 
 I 17.96 
 
 33.59 
 34.60 
 
 10.19 
 10.16 
 
 38.17 
 37.28 
 
 High protein .... 
 Low oil 
 
 Germs 
 Germs 
 
 \ 20.85 
 121.65 
 
 $21.70 
 
 34.99 
 36.02 
 
 25.01 
 
 10.12 
 10.07 
 
 13.13 
 
 34.04 
 32.26 
 
 40.16 
 
 High oil 
 
 Germs 
 
 (21.71 
 ( 17.55 
 
 24.62 
 41.76 
 
 13.36 
 
 8.75 
 
 40.31 
 31.94 
 
 
 
 I 17.84 
 
 41.75 
 
 8.81 
 
 31.60 
 
 Low protein 
 
 Endosperms 
 
 \ 5.69 
 ( 5.68 
 
 .83 
 .91 
 
 .43 
 .43 
 
 93.05 
 92.98 
 
 High protein .... 
 Low oil . 
 
 Endosperms 
 Endosperms 
 
 ( 13.67 
 1 13.92 
 
 { 9 13 
 
 .76 
 .72 
 
 52 
 
 .36 
 .41 
 
 .47 
 
 85.21 
 84.95 
 
 89.88 
 
 High oil 
 
 Endosperms 
 
 \ 9.14 
 ( 10.62 
 
 .51 
 1.07 
 
 .43 
 .36 
 
 89.92 
 
 87.95 
 
 
 
 1 10.10 
 
 1.24 
 
 .39 
 
 88.27 
 
 These results show in a very striking manner the effect of breeding in 
 changing the composition of the different physical parts of the kernel. 
 Thus the germs from the low oil corn contain about 25 percent of oil, 
 while those from the high oil corn contain nearly 42 percent of oil. As 
 
106 BULLETIN No. 87. [August, 
 
 stated above, breeding to change the oil content not only changes the 
 percentage of germ, but it also changes the percentage of oil in the germ. 
 It should also be noted that endosperms from the high oil corn contain 
 more than twice as much oil as those from the low oil corn, although the 
 percentage of oil in the endosperm is very small even in the high oil corn, 
 and this oil is largely contained in the horny gluten. 
 
 Perhaps the most marked and valuable results are shown in the per- 
 centages of protein contained in the endosperms from low protein and 
 high protein corn; the endosperms from the low protein corn contain less 
 than 6 percent of protein, while those from the high protein corn con- 
 tain nearly 14 percent of protein. These results, in connection with 
 others which we have given, would seem to prove very conclusively 
 that to select high protein seed corn by mechanical examination we should 
 select principally for a large proportion of the more nitrogenous part of 
 the endosperm; that is, the horny part. To select only for large germs 
 will have only a slight effect upon the protein content of the corn, al- 
 though it will produce a rapid and marked increase in the oil content. 
 
 Referring again to Table 20, it will be seen that the endosperms from 
 the high oil corn contain about one percent more protein than those from 
 the low oil corn. On the other hand, the germs from high oil corn con- 
 tain less protein (17.7 percent) than those from low oil corn (21.7 per- 
 cent), the difference being 4 percent protein in favor of the low oil corn. 
 
 These results were to be expected even from a study of the analyses of 
 the 163 ears reported in Bulletin No. 55, in 1899, which showed that large 
 germs were naturally even richer in oil than the size of the germs would 
 indicate, and that there is but very slight correlation between oil and pro- 
 tein, the increased oil tending to decrease the percentage, though not the 
 actual amount of protein in the germ. It will be seen from Tables 19 and 
 20 that the high oil corn contains nearly twice as much germ as the low 
 oil corn, and that the germs from high oil corn are more than one and one- 
 half times richer in oil than the germs from the low oil corn; but that, 
 although the high oil germs contain a larger total amount of protein 
 because of their increased size), they are really considerably poorer in 
 percentage of protein than the low oil germs. 
 
 It is perhaps worth while to consider the evident fact that, even if the 
 protein should increase in the germ in the same proportion as the oil 
 (which is not the case), we should need to increase the oil two pounds for 
 every one pound increase in protein obtained, if we depend upon the 
 method of picking our high protein seed corn by selecting for large germs. 
 In other words, to increase the protein in corn from 10 percent to 15 per- 
 cent by this method, would require the oil in the corn to be increased 
 from 5 percent to 15 percent. 
 
 Although we do not assume to say what should be the percentage of 
 oil in corn for feeding purposes, we do take the liberty of raising the ques- 
 
1903.] 
 
 THE STRUCTURE OF THE CORN KERNEL. 
 
 107 
 
 tion whether the popular opinion that the oil in corn should be increased 
 for feeding purposes may not be erroneous. Certainly the investigations 
 of Lehmann in Germany and of Shutt in Canada have indicated very 
 strongly that corn is already too rich in oil to be suitable as a foodstuff 
 for bacon hogs. It may also be called to mind that some other excellent 
 foodstuffs, such as oats, bran, barley, red clover and alfalfa, contain less 
 than half as much oil as is already contained in ordinary corn. 
 Attention is called to the fact that although the physical parts of the 
 corn kernel which contain nearly all of the oil (namely the germ and the 
 horny gluten) also contain most of the ash, yet a high percentage of ash 
 in the germs is associated with a low percentage of oil, and vice versa, 
 indicating that the ash content of the germ (which includes the major 
 part of the ash of the entire kernel) bears a more constant relation to the 
 oil-free material in the germ than to the whole germ. By computation 
 we find that the oil-free germs contain the percentages of ash given in 
 Table 21 (assuming the oil to contain no ash, which is approximately 
 correct*). 
 
 TABLE 21. PERCENTAGE OF ASH IN GERMS. 
 
 
 In fresh 
 germs. 
 
 In oil-free 
 germs. 
 
 From low protein corn . . 
 
 U0.19 
 
 15.34 
 
 From high protein corn 
 
 \ 10.16 
 ( 10.12 
 
 15.54 
 15.57 
 
 From low oil corn . 
 
 { 10.07 
 ( 13.13 
 
 15.74 
 17.51 
 
 From high oil corn 
 
 I 13.36 
 ( 8.75 
 
 17.72 
 15.02 
 
 
 { 8.81 
 
 15.12 
 
 Breeding for high or low protein produces no marked effect upon the 
 ash content or the oil content of either the germs or the endosperms, and 
 only slightly influences the protein content of the germs. (The low pro- 
 tein germs contain about 18 percent of protein and the high protein 
 germs about 21 percent.) The results show that such breeding produces 
 exceedingly marked effects upon the protein content of the endosperms, 
 the low protein endosperms containing less than 6 percent and the high 
 protein endosperms nearly 14 percent of protein. In this connection it 
 is well to remember that the corn kernel usually contains only about 11 
 percent of germ, while the endosperm amounts to about 89 percent of the 
 kernel. The significance of this becomes more readily apparent by an 
 examination of Table 22, which shows where the protein actually exists 
 in 100 pounds of corn. 
 
 *Actual determinations of the ash in corn oil have shown that the oil contains 
 only 0.2 per cent of ash. 
 
108 
 
 BULLETIN No. 87. 
 
 [August, 
 
 TABLE 22. PROTEIN IN 100 POUNDS OF CORN. 
 
 Names 
 of 
 parts. 
 
 Low protein corn. 
 
 High protein corn. 
 
 Differ- 
 ence. 
 
 Percent 
 of 
 corn. 
 
 Percent 
 of 
 protein. 
 
 Pounds 
 of 
 protein. 
 
 Percent 
 of 
 corn. 
 
 Percent 
 of 
 protein. 
 
 Pounds 
 of 
 protein. 
 
 Pounds 
 of 
 protein. 
 
 In germs . 
 
 9.33 
 90.67 
 
 18.01 
 5.69 
 
 1.68 
 5.16 
 
 11.44 
 88.56 
 
 21.25 
 13.80 
 
 2.43 
 12 22 
 
 .75 
 
 7.06 
 
 In endosperms 
 
 We thus find as a result of corn breeding that in the seventh genera- 
 tion we have a maximum difference of only .75 pound of protein in the 
 germs from 100 pounds of low protein and high protein corn, while in the 
 endosperms from these two kinds of corn we have a difference of 7.06 
 pounds of protein, in 100 pounds of corn. In other words, in changing 
 the protein content of corn the effect produced in the endosperms amounts 
 to almost ten times the effect produced in the germs. 
 
 THE COMPOSITION OF HOMINY MILL PRODUCTS. 
 
 Besides the investigations which we have carried on along this particu- 
 lar line in connection with our work of corn breeding, we have also made 
 some study of factory products, especially the products from hominy 
 mills, which make use of immense quantities of corn. 
 
 In the regular process of milling corn, a very large number of separa- 
 tions are made and several distinctly different final products are obtained, 
 some of which are composed almost entirely of certain distinct physical 
 parts of the corn kernel. The following is a brief and very general de- 
 scription of the usual process of corn milling : 
 
 The whole corn is somewhat softened by steaming and is then run 
 through a hulling machine, which not only removes the hull, but loosens 
 the germ and breaks off the horny gluten and more or less white starch. 
 The dust or pulverized material coming from the hulling machine con- 
 sists largely of white starch and horny gluten. The hulls and germs are 
 each separated but not in very pure condition, leaving what is termed 
 hominy, which consists chiefly of the horny starchy part of the kernel, 
 with more or less adhering white starch. 
 
 The product which is known as grits is made from the hominy and 
 consists of the horny starchy part separated in very pure form. In 
 making grits the coarse hominy is run through a grinding machine and 
 reduced to a coarse powder which may be termed coarse grits, much of 
 the adhering white starch being rubbed off from the horny starch in this 
 process. The coarse grits are then run through one or two more grinding 
 machines, until the horny starch is reduced to a rather fine powder, which 
 may be termed fine grits. This material consists of the horny starch in 
 very pure condition. After each grinding the fine dust consisting largely 
 
1903.] 
 
 THE STRUCTURE OF THE CORN KERNEL. 
 
 109 
 
 of the white starch is separated from the grits and goes into the product 
 known as corn flour. 
 
 . In addition to the corn flour thus regularly separated and handled in 
 considerable quantities, there is constantly produced a small amount of 
 what is termed " break" flour. This is an exceedingly fine dust also pro- 
 duced in the process of breaking the corn particles in the grinding machines 
 which reduce the hominy to grits. The break flour is carried from the 
 machine by an air current through conduits and finally collected. This 
 is another very pure form of the white starch. 
 
 Thus, in the regular milling process there are two physical parts of the 
 corn kernel separated in very pure form: namely, horny starch (fine 
 grits) and white starch (break flour or corn flour) and two other distinct 
 parts which are separated somewhat less perfectly, the hulls and the 
 germs. 
 
 By the courtesy of the manager of the American Hominy Company's 
 Mills at Decatur we were allowed to collect representative samples of 
 these different products for analysis. The composition of these products 
 is given in Table 23, and it will be found interesting to compare these 
 results with the composition of the same products, or parts, which were 
 obtained by exact hand separation, as given in Table 1. For conve- 
 nience in comparison Table 23 also shows the composition of these parts 
 as obtained from Ear No. 1 (Table 1), which is very similar to the corn 
 which was being used in the mill at the time the samples were taken. 
 This was fairly representative of the ordinary white corn grown in 1902, 
 nearly all of which was abnormally low in protein, owing to seasonal in- 
 fluences. 
 
 TABLE 23. COMPOSITION OF PARTS OP THE CORN KERNEL SEPARATED BY 
 HOMINY MILL AND BY HAND. 
 
 Names of parts. 
 
 Methods of 
 separation. 
 
 Protein, 
 percent. 
 
 Oil, 
 
 percent. 
 
 Ash, 
 percent. 
 
 Carbo- 
 hydrates, 
 percent. 
 
 Hulls .. 
 
 By mill 
 
 6 85 
 
 2.94 
 
 1.11 
 
 89.10 
 
 Hulls 
 
 By hand 
 
 4 97 
 
 .92 
 
 .82 
 
 93.29 
 
 Horny starch (fine grits) . . 
 Horny starch 
 
 By mill 
 By hand 
 
 8.46 
 8.12 
 
 .44 
 .16 
 
 .26 
 .18 
 
 90.84 
 91.54 
 
 White starch (corn flour*) . 
 White starch (break flour) 
 White starch (from tip) . . . 
 
 Germs 
 
 By mill 
 By mill 
 By hand 
 
 By mill 
 
 5.91 
 5.88 
 6.10 
 
 15.84 
 
 1.63 
 2.04 
 .29 
 
 21.26 
 
 .49 
 .68 
 .29 
 
 7.41 
 
 91.97 
 91.40 
 93.31 
 
 55.49 
 
 
 By hand 
 
 19 91 
 
 36 54 
 
 10 48 
 
 33 07 
 
 Whole corn 
 
 Mill sample 
 
 9 31 
 
 4 20 
 
 1.43 
 
 85.06 
 
 Whole corn 
 
 Ear No. 1 
 
 9.28 
 
 4.20 
 
 1.41 
 
 85.11 
 
 *Obtained by direct separation from grits. 
 
110 BULLETIN No. 87. [August, 
 
 In general the composition of these mill separations agrees with the 
 composition of the same parts separated by hand, although in nearly 
 all cases the mill products show more or less contamination or mixture 
 with other parts of the kernel. Thus the mill hulls are noticeably high 
 in protein and oil owing to the presence of some particles of horny gluten 
 and germ; while the mill germs are too low in protein and oil because of 
 the presence of some hulls and tip caps. Furthermore, some oil is lost 
 from the germ and absorbed by other parts in the milling process. 
 The fine grits are almost pure horny starch, except that they contain 
 about twice as much oil as the hand-separated product. This is doubt- 
 less due to the fact that some germs are broken or crushed in the hulling 
 machine and the liberated oil is absorbed to some extent by the hominy, 
 chiefly, of course, by the white starch, as indicated by the high oil content 
 of the break flour and the other regularly separated corn flour, although 
 it is evident that a small portion of this liberated oil remains adhering to 
 the fine grits. The white starch contains 5.88 to 5.91 percent of protein, 
 while the horny starch (fine grits) contains 8.46 percent, or almost one- 
 half more. 
 
 It will be observed that the two samples of whole corn are almost 
 identical in composition. While the corn is fairly representative of 
 much of the white corn grown during the season of 1902, attention is 
 called to the apparent fact that this is not the most suitable corn for the 
 manufacture of hominy and grits. It seems evident that corn containing 
 a higher percentage of the horny starchy part would be more valuable for 
 the hominy mill. The manager of the American Hominy Company's 
 Mills at Decatur has assured the writer that he prefers corn which shall 
 run high in grits (horny starch), but he does not desire that the oil content 
 should be increased; indeed, it would be much better for milling purposes 
 to have the percentage of oil in corn reduced, because of the difficulty of 
 preventing the oil from being absorbed by other products and injuring 
 their quality, the tendency being for the oil to become rancid when ex- 
 posed to the air. The hominy mills offer some encouragement to farmers 
 to grow corn especially suited to their use. 
 
 SUMMARY OF BULLETIN No. 87. 
 
 The investigations reported in this bulletin serve to establish the fol- 
 lowing facts: 
 
 1. The kernel of corn consists of six readily observable and distinctly 
 different physical parts, which are known as (1) the tip cap, (2) the hull 
 (3) the horny gluten, (4) the horny starch, (5) the white starch, (6) the 
 germ. 
 
 2. The tip cap covers the tip or base of the kernel and comprises only 
 about 1.5 percent of the grain. 
 
 3. The hull is the very thin outer coat. It comprises about 6 percent 
 
190:5.] THE STRUCTURE OK THE CORN KEHNEL. Ill 
 
 of the kernel and contains a lower percentage of protein (about 4 percent) 
 than any other part of the kernel. 
 
 4. The horny glutenous part (aleurone layer) lies underneath the hull 
 surrounding the kernel. It comprises from 8 to 14 percent of the grain 
 (being more abundant in high protein corn), and it contains from 20 to 
 25 percent of protein, being the richest in protein of all the parts of the 
 corn kernel. 
 
 5. The horny starchy part is the chief substance in the sides and back 
 of the kernel (the germ face being considered the front of the kernel). 
 This substance comprises about 45 percent of ordinary corn, but is much 
 more abundant in high protein corn and less abundant in low protein. 
 Although rich in starch it contains about 10 percent of protein (more in 
 the high protein corn and less in low protein corn). It contains a greater 
 total amount of protein than any other part of the kernel. 
 
 6. The white starchy part occupies the center of the crown end of the 
 kernel and usually partially surrounds the germ. It comprises about 25 
 percent of the kernel (less in high protein corn and more in low protein 
 corn). It is poor in protein (5 to 8 percent). 
 
 7. The germ occupies the central part of the kernel toward the tip 
 end. It comprises about 11 per cent of the kernel (more in high oil corn 
 and less in low oil corn). The germ contains from 35 to 40 percent of 
 corn oil or from 80 to 85 percent of the total oil content of the corn kernel. 
 
 8. High protein corn contains a large proportion of the horny parts 
 (both of the horny glutenous part and the horny starchy part), and a 
 correspondingly smaller proportion of the white starchy part. The horny 
 parts comprise more than 60 percent of high protein corn and contain 
 about 80 percent of the total protein content of very high protein corn. 
 
 9. The value and reliability of the method proposed in previous 
 bulletins by which any farmer can select high protein seed corn (selecting 
 for a large proportion of horny parts) by a simple mechanical examina- 
 tion of the corn kernels has been fully confirmed by the results which 
 have been subsequently obtained and which are now reported in this 
 bulletin. 
 
 10. The value of the method proposed for picking out high oil seed 
 corn by selecting for a large proportion of germ is also fully established. 
 
 11. The degree of correlation existing in the corn kernel between the 
 percentages of germ and protein is very slight and is frequently entirely 
 absent, consequently the proportion of germ in the corn kernel is not a 
 reliable index of its protein content. 
 
 12. The composition of the different products obtained from corn by 
 hominy mills, as well as by other factories, serves greatly to emphasize 
 the importance of breeding corn for special purposes. 
 
 (For more complete details of corn breeding, the reader is referred to 
 Bulletin .No- 82. " Methods of Corn Breeding," which can be obtained upon 
 
112 BULLETIN No. 87. [August, 1903. 
 
 request from the Illinois Experiment Station, Urbana, Illinois. The 
 supply of Bulletin No. 53 and Bulletin No. 55, both relating to corn 
 investigations, is now exhausted.) 
 
UNIVERSITY OF ILLINOIS-URBANA