63 



which shows an apparent negative influence in the "normal" column 

 and a positive influence of ten bushels per acre annually in the col- 

 umn based on the average check. The apparent negative influence 

 is due directly to extra yields in the adjacent check plot 7 which is 

 apparently receiving some benefit from plot 8 as a result of cross- 

 feeding. The same thing appears in the negative figures shown by 

 plots II and 12, though they are not entirely eliminated by using the 

 average check as the basis. The average check, however, is not 

 entirely free from the cross-feeding influences, since it only dis- 

 tributes the extra yields and hence it is probable that the negative 

 figures in plots 11, 12, and 15 of the last column merely indicate 

 that these materials have no definite influence when applied by 

 themselves, and the further deficits are attributable simply to natural 

 fluctuations. 



Returning now to the results themselves, and especially to those 

 treatments not included in the preceding experiments, we may note 

 first that the muriate of potash in plot 5 has given much better 

 gains than the sulphate in the adjacent plot. This is contrary to the 

 results of the Massachusetts Experiment, but similar results are 

 now being shown in all of our own experiments wherever this com- 

 parison occurs. Hence the differences in the Massachusetts experi- 

 ment would seem to be due to something other than the difterence in 

 potash carriers. At present, therefore, we believe that the muriate 

 is at least as efficient as the sulphate and in view of the facts that it 

 is cheaper, more soluble, and much less subject to "caking" in the 

 mixtures, we are now using and recommending it for apples. 



In plots II and 12, and other similarly treated plots in our ex- 

 periments, we see the apparent futility of attempting to materially 

 improve yields by applying phosphates alone. This is not due to the 

 fact that phosphates are not needed, nor can it be largely attributable 

 to the absence of cultivation, as may be seen by comparing the re- 

 sults in plots 9 and 2. Nitrogen, by itself in No. 9, shows an annual 

 gain of 96 per cent, or 113 bushels per acre, but when phosphorus 

 is added in plot 2, these benefits are more than doubled. Phos- 

 phorus as usual, therefore, appears to be next in importance after 

 nitrogen in improving yields. 



The Permanence of Fertilizer Influence. — It is another fairly 

 common impression that the influence of fertilizers is transient and 

 that, even where their effect is favorable at first, this effect soon 

 wears out and may leave the soil worse than before. This evi- 

 dently depends very largely on the character of the fertilization and 

 in this respect apples are not different from other crops. If the 

 gains are induced by some caustic action of such materials as 

 gypsum or lime when used alone, this may actually be the final 

 result. 



On the other hand, it should be noted here that in plots 2, 3 

 and 8, where definite plant foods are being supplied, the effects of 

 fertilization were greater than ever before in 1912, the sixth year 

 of the experiment. The steadiness and regularity of the increases, 



