THE BIOLOGICAL METHOD OF DETERMINING THE FERTILIZER 

 REQUIREJMENT OF A PARTICULAR SOIL OR CROP. 



E. V. HIBBARD AND S. GERSHBERG. 



Soil experts and chemists are loatli to put any dependence on tlie result of 

 a chemical analysis of either the crop or the soil as far as using: them to plan 

 a fertilizer treatment. The best and only logical way, then, to determine the 

 crop's needs is to grow the crop on the soil and make a detailed study of the 

 relations between the two. The great majority of fertilizer experiments have 

 not been so planned. The combinations or ratios of the three fertilizer con- 

 stituents commonly used have been greatly restricted, and except in a few 

 cases the soil has been ignored. 



The method suggested here consists in a well planned series of combina- 

 tions, each one vai-ying from the other and all the rest in small increments. 

 In this way a wide range of ratios or combinations are used and thei'e are 

 no missing ones in the whole sei'ies. Obviously this is a more logical method 

 than the hit-or-miss one, in which the combinations are selected at random. 



To indicate the method by which a series of combinations is planned, since 

 space for a figure is not available, one must picture to himself an equilateral 

 triangle, the kind that lias often been used when one desires to express in a 

 diagrammatic manner the relations between three variables or three quantities, 

 such as the three fertilizer salts, acid phosphate, sodium nitrate and potassium 

 sulphate. This particular triangle which we wish you to think of indicates 

 at the points of intersection of the various lines all the possible combinations 

 or ratios of the three salts, where each constituent varies from the other by 

 increments of ten per cent. There are eight horizontal rows of combinations, 

 counting frota the base to the apex of the triangle. In the lower row there are 

 eight combinations ; in the one above, seven, and in each succeeding row 

 one less. 



Many such series or sets of combinations have been arranged for water 

 cultures in the greenhouse. This same method can be used in the field, but 

 when the plots are large, as for example one-twentieth of an acre, the work 

 would be difficult and almost impossible. However, the plots can be cut down 

 to small sizes and a number of checks or controls introduced to average up 

 individual differences. Then at the end of the experiment the undesirable 

 combinations can be eliminated and the others run on plots of larger size. 

 Another way to obviate this difficulty would be to run a series with few num- 

 bers of combinations. Let the combinations difl'er in increments of 20 per cent. 



In the summer of 1918 a complete series of 36 combinations, run in tripli- 

 cate, was put out on the station farm. The experimental plots were 2 feet by 40 

 feet. Besides the 36 combinations of three fertilizer salts, there were several 



21st Mich. Acad. Sci. Kept. 



