kernel by selection and breeding. At that time the experiments had been under way only about two years, but already results had been obtained sufficiently pronounced to indicate the possibility of changing the composition of the corn kernel as regards protein, oil and carbohydrates. I n fact, in his conclusion the author remarks that “while i t may require ten or twenty years’ work to enable one to form a very definite opinion as to the extent to which it C possible to influence the chemical composition of corn, it is believed that the data thus far obtained may be of practical and scientific value.” Since the time this statement was made this investigation has been continued and twelve years’ work has been added to these experiments. N o T A B L E I.-PERCENTAGE O F MnO FOUND B Y W A L T E R S ’ L f E T H O D A S DEfurther report, however, has been given to this Society, SCRIBED BY WASHINGTON, AND BY OUR NEW.METHOD. txp. Soil Walters’ method, MnO in extractions I-IV. Authors’ and in view of this fact i t was thought that the members No. No. I. 11. 111. 1x7. Total. method. would be interested to learn regarding the present 1 3490 0,082 0.020 0.004 trace 0.106 0.109 status of the work. 2 3490 0.080 0.027 0.005 ? 0.112 0.111 3 3490 0,060 0.042 trace ? 0.102 Inasmuch as the general purpose of the investiga4 3490 0,095 0.013 trace ? 0.108 tion and the methods of carrying i t out have been 5 3490 0,095 0.013 trace ? 0.108 given already quite fully in the above-mentioned 6 3490 0.064 0.016 not 7 3490 0.046 0.029) determined paper and elsewhere (see Bull. 128, Illinois Agri8 3486 0.045 not determined 0.107 cultural Experiment Station), it will be the present 9 3486 0.052 0.105 10 3487 0.018 0.111 plan t o mention only as much of this matter as is 11 3487 0.028 0.110 necessary to a n intelligent understanding of the re12 3488 0.020 0,111 sults, the chief purpose being to give a brief resume 13 3488 0.112 14 3489 0.025 0.120 of the results up-to-date. 1.5 3489 0.031 0.119 Stated briefly, the proposition involved in this 16 3491 0.044 0.123 17 3491 0.044 n.125 investigation was to take a variety of corn and through 18 3483 0.060 0,096 selection and breeding try to influence the composition 19 3483 0.060 n 098 of the grain in four different directions, namely: I , 20 3459 0 033 n 067 21 3459 0.067 Increase the protein content; 2 , Decrease the protein StJMMA R S. content; 3, Increase the oil content; 4,Decrease the oil I . We have found that Walters’ method for the content. determination of small quantities of manganese, as Through a preliminary study involving the analyses described by Washington and Hillebrand, is not of a good many samples of different ears and parts applicable to Nebrasks soils, but that in some instances of ears the principles of the selection were established. four extractions on a one-gram sample with hydro- I t was found that there is considerable variation fluoric and sulfuric acids were necessary before all among different ears with respect to composition, of the manganese was obtained in solution. while the kernels within a single ear are comparatively 2 . We have found that when such soils are fused uniform; that is to say, the ear can be taken as the with sodium carbonate and the solution of this melt unit of selection and this gives the working basis. is acidified with sulfuric acid and oxidized by means ‘rhus the method is to take two or three rows of of sodium bismuthate, that all of the manganese is kernels lengthwise of the ear as a sample for analysis. oxidized to permanganic acid, and may easily be If the ear proves to be suitable for seed the remaining estimated colorimetrically. kernels are planted. 3. We claim for this method greater rapidity, Starting in 1896 with a variety then known as absence from irritating fumes, less danger of loss by Burr’s White, 163 ears were chosen and analyzed, spattering, more ease of operation in unexperienced determinations being made for protein, oil, ash and hands, duplicate determinations on one samplr of carbohydrates. According to the percentage of thc soil, and much greater accuracy. constituents, four lots of these ears were picked out with which to start the four respective breeding plots, ALTERING THE COMPOSITION OF INDIAN CORN BY SEED a high protein plot, a low protein plot, a high oil plot, and a low oil plot. Each ear was planted in a SELECTION.’ separate row, the different plots being isolated in By L. K. S M I T H . order to avoid contamination from cross pollination. Received January 8 , 1912. I n a paper presented before the American Chemical From the harvest a large number of ears were analyzed Society in 1899 Dr. C. G. Hopkins reported a line from each plot, and based upon these results ears of experiments undertaken a t the Illinois Experiment were selected for seed with which t o plant the respectStation t o influence the composition of the corn ive plots the following year. This general plan, with a few minnr changes of 1 Presented at A. C. S. Meeting at Washington. D C . , December. 191 I .
is calculated. After a little practice, the complete operation can be carried out in an hour and a half, so t h a t in this respect alone the method is greatly to be preferred. When duplicate determinations are desired only one fusion is necessary, for two separate oxidations may be taken from the total 250 cc. I n the data which are given in the last column of Table I each figure represents a separate fusion and oxidation, and in the sixteen determinations, representing duplicate fusions on eight different soils, the maximum difference between duplicates amount to only 0 . 0 0 2 per cent. N o determinations were discarded in compiling this table.
July, 1912
T H E J O U R N A L OF I N D C J S T R I A L AA’D E N G I N E E R I N G C H E i V I I S T R Y .
detail, has been followed year after year until we now have the results for 14 generations (the samples for the fifteenth generation have not yet been analyzed). In this manner four different strains emanating from the same variety have been derived. I t has been extremely interesting to watch their gradual evolution. This may be shown in outline by the following tabular arrangement : :‘
Year. 1896 1897 1898 1899 1900 ‘1’901 1902 1903 1904 1905 1906 1907 1908 1909 1910
Year. 1896 1897 1898 1899 1900 1901 1902 1903 1904 1905 1906 1907 1908 1909 1910
BREEDING FOR HIGH AND LOW PROTEIN. Average per cent. protein in crop each generation. High protein. Low protein. Difference. 10.92 11.10 11 .05 11.46 12.32 14.12 12.34 13.04 15.03 14 72 14.26 13 89 13 94 13.41 14.87
10.92 10.55 10.55 9.86 9.34 10.04 8.22 8.62 9.27 8.57 8.64 7.32 8.96 7.65 8.25
0.00 0.55 0.50 1.60 2.98 4.08 4.12 4.42 5.76 6.15 5.62 6.57 4.98 5.76 6.62
BREEDING FOR HIGHA N D L O W OIL. Average per cent. oil in crop each generation. High oil. Low oil. Difference. 4.70 4.73 5.15 5.64 6.12 6.09 6.41 6.50 6.97 7.29 7.37 7.43 7.19 7.05 7.72
4.70 4.06 3.99 3.82 3.57 3.43 3.02 2.97 2.89 2.58 2.66 2.59 2.39 2.35 2.11
0.00 0.67 1.16 1.82 2.55 2.66 3.39 .3.53 .%.08 4.71 4.71 4.84 4.80 4.70 5.61
525
now nearly twice as rich in protein as the other, and two other strains have been developed, the one of which is now more than three times as rich in oil as the other. The question arises as to the effect of this selection upon the composition of parts of the plant outside of the kernel. This matter has been studied by analyzing each year different plant parts in the various strains. Without entering into any discussion of the results i t may be stated in a general way that as yet we have been able t o find no pronounced changes in the composition of other parts of the plant resulting from the changes produced in the grain. Another practical question always arises in this connection, and that is regarding the effect of this selection upon productive qualities. I t may be said that in spite of the rigorous selection for special characters to which this corn has been subjected, the yields in these four strains have, in general, been quite well maintained. Selection for high protein seems to have been accompanied in this instance by a somewhat diminished yield as measured by bushels of corn per acre. If, however, we place it upon the basis of pounds of protein per acre our high protein strain is sufficiently higher in percentage, so that it yields more protein per acre than the ordinary varieties growing alongside i t in the productivity tests. AGRICULTURE. UNIVERSITY OF ILLINOIS.
COLLEGE O F
RELATION OF ACTIVE POTASH TO POT EXPERIMENTS.’ B y G. S. FRAPS. Received .March 25. 1912
By the term “active potash” we mean that soluble in N / s nitric acid. This solvent has been in use for some time for extracting the more soluble potash and phosphoric acid from the soil. As pointed out in a In considering these data attention should be called previous article, the quantity of phosphoric acid to the fact that environmental conditions exert an extracted is related to the results of pot .experiments influence upon the protein as well as the oil content on the soil. Similar experiments will here be reported and a study of these results reveals some pronounced on potash. Pot experiments with representative Texas soils seasonal effects. By having a “high” and a “low” plot of each, however, we obtain a check which enables have been made at this station for seven years. The us to eliminate these seasonal effects t o some extent. experiments have been carried on under diverse conThus the last column of the table showing the dif- ditions, sometimes not favorable tothe bestgrowthof the ference in per cent. each year furnishes the best plants. The plants were grown in Wagner pots containing 500 grams soil, applications of fertilizer being made information of the real progress due t o heredity. Without going furt,her into a detailed discussion to the various pots. The pots were kept in houses of the figures we may say by way of summary that covered partly or completely with canvas, in order starting with a variety of corn of average composition to moderate the temperature. A crop is regarded as i t has been possible by 14 years of breeding to very deficient if it is only 50 per cent. or less, of the increase ,the protein content from 10.92 per cent. completely fertilized crop. If less than 90 per cent., in the original, t o 14.87per cent. in the crop of 1910, i t is considered deficient (D). If between 90 and I I O at the same time by selection in the opposite direction, per cent., i t is considered as not deficient (S). If to decrease this constituent to 8 . 2 5 per cent. Even over I I O per cent., i t is considered that the fertilizer more striking changes than these have been produced injures the crop (T). The crop without potash (PN) in the oil content. The oil has been increased from is always compared with the crop with potash ( P N K ) . 4.70 per cent. in the original to 7 . 7 2 per cent in the Relation of Deficiency to Active Potash.-The 172 crop of 1910 while a t the same time in another strain soils studied have been divided into groups according i t has been decreased to 2 . I I per cent. to thcir content of active potash. The following I n other words, out of a single variety of corn 1 A full account of this work will he published as Riillefzn 146 of thr two strains have been developed of which one is Texas Experiment Station.
