45 



to an exceptional crop tliat occurred on this plot in 1909, and from 

 which the plot has not yet recovered. In the next plot, we see the 

 eltect of adding a mulch to the sod treatment. In this case, al- 

 though all the herbage that grows is left in the orchard, and a 

 further application of 3 tons of straw per acre is added to the plot, 

 the average annual gain is only 22 bushels per acre. In the next 

 plot, we find that tillage and leguminous cover crops have given a 

 fair increase, amounting to 100 bushels per acre on the average. 

 This, however, is hardh" to he compared with the 452 bushel in- 

 crease shown in the next case, which is obtained without tillage 

 of any kind, merely by the addition of a fertilizer that carries the 

 elements that are evidently lacking. 



In some quarters one would gather the impression that apples 

 can scarceh' be grown without tillage. AMiile we have nothing 

 against proper tillage as an orchard treatment, yet this and other 

 results from our experiments show that it is by no means indispen- 

 sible in the production of first grade apples and that it can be 

 readily over-emphasized like anything else. There are many situa- 

 tions that are otherwise very well suited for apples, where tillage 

 is decidedly inadvisable, and where, with proper management, the 

 trees would get along very much better without it. In such situa- 

 tions it is undoubtedly preferable to sow the orchard down to some 

 leguminous crop as a permanent cover and follow the mulch system, 

 properly supplementing it with fertilization. For this purpose, 

 hairy vetch is doubtless preferable, on account of its relatively low 

 moisture draft, and its usually excellent staying powers when once 

 well seeded down. Whenever it is crowded out by the grasses, the 

 orchard may be re-plowed and again sowed to vetch, if the trees 

 seem to require it. 



Data on Fertilizers from Other Experiments. 



Thus far we have confined our attention to a single experiment, 

 primarily because the contrasts in it are so great that both the exist- 

 ence and nature of the effects could scarcely fail to be recognized. 

 To go through each experiment in this way would be impossible 

 in our present space, hence we have condensed into the next two 

 tables a statement derived from the results of six experiments, in- 

 cluding the one just discussed. These tables show t^lie average ef- 

 fects of the dift"erent fertilizer elements, obtained in six experi- 

 ments, during periods covering from three to five years as indicated. 

 The eft'ects are calculated as closely as possible and are expressed 

 in terms of per cents, of benefit based on the normal performance 

 of the treated plots. The methods followed in making the calcu- 

 lations are described briefly in our Bulletin 100 from the Pennsyl- 

 vania Station, and described in full in our Annual Report for 19 To- 

 il. 



