334 



SOILS. 



taken out of the soil by one crop are in general of the same 

 order of magnitude as the above, what is taken out by one or 

 two crops will usually fall within the limits of analytical errors, 

 especially of those incurred in sampling the soil. Yet that the 

 changes caused by a number of successive crops can be proved, 

 even by the ordinary methods, has been abundantly verified. 

 For it seems that the losses of soil ingredients in cultivated 

 lands exceed considerably those calculated from the actual 

 drain represented by the crops. 



Plot Tests. There is, however, an obvious and apparently 

 simple method by which every farmer might make his own 

 fertilizer tests, on a small and inexpensive scale, the results of 

 which may afterwards be put in effect on his entire land. It 

 is to apply in proper proportions on plots (of say from one 

 twentieth to one fortieth of an acre), the several plant-food 

 ingredients usually supplied in fertilizers, singly as well as 

 conjointly w r ith each other, leaving check unfertilized plots 

 around as \vell as among them. By comparison with these, 

 the cultural results should at once determine which of the 

 fertilizers can most advantageously be applied to the land. 

 Such tests when carried out with all the proper precautions 

 are often very decisive and practically successful. But they so 

 frequently suffer from seasonal influences (such as scanty or 

 excessive rainfall, cold or heat, etc.), inequality of soil condi- 

 tions, failure to apply the fertilizers at the right time, or in the 

 right way, the depredations of insects and birds, and other 

 causes, that it generally takes several seasons' trial to obtain 

 any definite results. On level lands of uniform nature and 

 depth, they are most likely to be successful ; while on undu- 

 lating or hill lands it is not only very difficult to secure uni- 

 formity of soil and subsoil on areas of sufficient size, but also 

 to prevent the washing of fertilized soil, or fertilizer in solu- 

 tion, from one plot to the other by the influence of heavy rains 

 or irrigation; thus wholly vitiating the experiments. In very 

 many cases, especially in the arid region, the results of such 

 trials have been practically nil, for the reason that physical de- 

 fects of the soil, and not lack of plant-food, were the cause of 

 unsatisfactory production. 



A full examination of physical conditions, as outlined in 

 previous chapters, should in all cases precede the application of 



