146 Bulletin 129. 



size of the plats can be reduced to one-twentieth of an acre ; if 

 the field is fairly uniform in character, the results with smaller 

 plats might be just as reliable as those with larger plats. Oi 

 course only half as much of each fertilizer will be required on 

 plats half as large as those indicated in the diagram. 



The growing crops should be looked over from time to time, 

 and a record should be kept of any diflferences observed between 

 the different plats, of attacks of insect pests or of fungous dis- 

 eases. A record of the weather should be kept. 



The harvesting of the crop.— In carrying out this part of the 

 work, allowance must be made for the possible growth of the roots 

 of one row into the feeding- ground of the adjoining rows ; thus 

 the outside row of one plat may steal food from the next plat that 

 was not intended for it ; hence the directions to exclude the two 

 outside rows of each plat, one on one side and the other on the 

 other side, and not to include the crop of those rows in the har- 

 vest measured, are important. Since this cannot be done conven- 

 iently with wheat and other small grains, it is recommended to 

 leave vacant spaces of two feet between the plats. 



In measuring the crop, due credit should be given for every 

 part of it that can be utilized in any way ; if corn, not only the 

 seed, but the stalks ; if wheat, oats, etc., the straw as well as the 

 seed ; if potatoes, of course only the tubers. 



The results. — In spite of all the care that may be taken in car- 

 rying out these soil tests, irregularities will appear in the results, 

 sometimes so great that it is not easy to decide what their real 

 meaning is ; weather, soil, tillage and fertilizer work together 

 for the making of the crop on each of the plats. All the plats 

 suffer alike in bad weather and rejoice alike in good weather ; the 

 tillage has been the same for all the plats ; all the plats were 

 planted on the same day ; only the fertilizing is diflferent, because 

 we made it so. The soil we suppose to be at least fairly uniform 

 throughout ; if, however, it does differ seriously from end to end 

 of the field, all our long and narrow plats run across these irregu- 

 larities and should be affected alike by them. 



There may, however, be differences not easy to detect in the 

 soil of diflferent plats, by which the growth on one plat may be 

 specially favored, or by means of which the fertilizer may be bet- 



