630 



SCIENCE 



[N. S. Vol. XXVII. No. 694 



the same sample of soil are recorded as differ- 

 ing by as much as 6.38 and 12.34 parts per 

 million for PO.; 3.15 and 23.39 for NO3; 9.43 

 and 15.33 for K. It would be strange indeed 

 if, under such conditions, anything but a 

 medley of results were secured. 



But the results of a whole season of much 

 more critical and reliable work along these 

 lines which preceded the gathering of the data 

 in Bulletin 22 and a second year's work fol- 

 lowing this, still more critical and exhaustive, 

 are left out of consideration wholly in reach- 

 ing the conclusion under discussion. The first 

 year's work referred to pointed strongly to the 

 conclusion that, where other conditions of 

 growth are equally favorable, the largest yields 

 of crops are not only associated with the soils 

 yielding the largest amount of water-soluble 

 salts, but that the sap of crops growing on 

 such soils is also richer in the same salts. In 

 other words, it had been shown that on soils 

 from which the largest amounts of water- 

 soluble salts could be recovered more of the 

 same salts were carried in the sap of the plants 

 growing upon these soils, and the data may be 

 found in Bulletin No. 26 of the Bureau of 

 Soils, in which the letter of transmittal 

 acknowledges the accuracy of the data but 

 rejects the conclusions drawn therefrom. 



The second year's work, whose results are 

 throughout in accord with those of the first 

 year, show the following relation between 

 yields and the water-soluble salts which had 

 been recovered from eight soil types upon 

 which two crops were grown to maturity under 

 normal field conditions and where the yields 

 were accurately determined by weighing. The 

 soil types used were selected with a view to 

 having, those of different productive capacity, 

 and four of those were rather above, and the 

 other four rather below, the average of good 

 soils. Taking the yields and the water-soluble 

 salts recovered from the poorer soils as 100, 

 the relations found are as stated below: 



Four Four 



Poor Good 



Soils Soils 



N-|-P-|-K in samples once washed. . 100 188.5 



N-l-P-f K in " 11 times " . . 100 193.1 



Yield of shelled com 100 191.7 



Yield of potatoes 100 271.4 



It is here clear that there is a marked in- 

 crease in yield, both of com and of potatoes, 

 where the water-soluble salts carrying nitro- 

 gen, phosphorus and potassium are present in 

 largest removable amounts, whether this is 

 shown by one or by eleven washings of the 

 same sample. 



Not only was this strongly marked relation 

 shown to exist when the soils are thus grouped, 

 but four equal areas of each of the eight soil 

 types were treated to 300 pounds of guano and 

 other fours to 5, 10 and 15 tons of stable 

 manure respectively, and the differences in re- 

 coverable water soluble salts determined 

 throughout the season, together with the yields 

 of corn and potatoes, with the results given 

 below. In this table each value is a mean 

 from 32 field plots. 



It is thus here shown that with each in- 

 crease in the amount of the three plant food 

 elements which could he recovered from the 

 soil there was a regular and corresponding in- 

 crease in the yield of each of the two crops, 

 the gain in the corn being almost directly pro- 

 portional to the increase of the three plant 

 food elements recovered by water from the soil. 

 With data thus strongly indicating larger 

 yields associated with increasing amounts of 

 water-soluble salts recovered from the soil, 

 collected before the data of Bulletin 22 and 

 confirmed by a second year of still more 

 critical work, the whole work is ignored and 

 the reverse relation held to be demonstrated by 

 data of the character stated. 



Soluble Salt Content in Soils Constantly 



Maintained. 



This is another of the contentions woven 

 into the fabric of Bulletin 22 but without 

 basis in recorded data other than that spread 

 upon its pages in the form of analyses of 

 some Eothamsted soils which are offered to 

 introduce the statement "that the decreasing 



